MOST GENITAL CHLAMYDIAL INFECTIONS IN females are asymptomatic,1 and therefore detection of infection often relies on diagnostic testing. Nucleic acid amplification tests (NAATs) are useful for chlamydial screening in asymptomatic females because they are highly sensitive and specific and can be performed on less invasive specimens (e.g., urine or self-obtained vaginal swabs), which provides an opportunity to screen for chlamydial infection in nontraditional settings (e.g., schools, correctional facilities, and so on) without performing a pelvic examination. A concern has been that screening in the absence of a pelvic examination will miss important clinical information.
Our understanding of the frequency and relationship of clinical signs to chlamydial infection is based mostly on prior chlamydial screening studies that used tests with a lower sensitivity than NAATs, primarily chlamydial cell culture. Such studies have reported that pelvic examination findings of mucopurulent cervicitis (MPC), which include mucopurulent endocervical discharge and easily induced endocervical bleeding on collection of endocervical swabs, are associated with chlamydial infection,2–7 and that MPC findings are not uncommon in asymptomatic females, often being present in more than one third of chlamydia-infected female patients.7,8 This introduces the concern of missed opportunities for early treatment of chlamydia in asymptomatic women undergoing chlamydial screening without a pelvic examination. Indications for empiric chlamydia treatment in asymptomatic females include pelvic examination findings of MPC or pelvic inflammatory disease (PID) or sexual contact with a partner with chlamydia. Early chlamydia treatment may limit further transmission and also the development of PID and its associated reproductive health complications (e.g., infertility).
Studies on the frequency of pelvic examination findings in asymptomatic females with chlamydia detected by NAATs are sparse. A recent study by Marrazzo et al assessed the impact of patient characteristics on NAAT performance and demonstrated mucopurulent endocervical discharge was present in over 27% of chlamydia-infected study participants and, when discharge was present, endocervical chlamydial polymerase chain reaction had a higher sensitivity; endocervical discharge was present in only 13% of chlamydia-negative females.9 However, their study included patients with genital symptoms, gonorrhea, and/or contact with chlamydia or gonorrhea and did not evaluate the MPC sign of easily induced endocervical bleeding.
The primary objective of our study was to assess the relationship of chlamydial infection with pelvic examination findings in asymptomatic females undergoing routine chlamydial screening by NAATs. We hoped that study findings would improve our understanding of the frequency of MPC and PID findings in young asymptomatic women tested by NAAT and hence what proportion of these women may not have received empiric chlamydia therapy if their screening had not included a pelvic examination. A secondary objective was to evaluate differences in frequency of pelvic examination findings in chlamydial NAAT-positive women based on chlamydial culture results. Because many prior studies evaluating pelvic findings and chlamydia used culture, which would be expected to identify infections with a higher chlamydial organism burden, we hypothesized that the frequency of examination findings would likely differ in culture-positive versus culture-negative individuals, with the latter having a lower frequency of examination findings as a result of the lower chlamydial organism burden.
Materials and Methods
Study Population and Data Collection
We retrospectively analyzed data collected in a multicenter cross-sectional study of chlamydia positivity in asymptomatic females screened with different NAATs comparing different genital specimen types.10 In this study, subjects were excluded if they had genital symptoms, antibiotics in the prior month, or symptomatic partners. All study participants were tested for gonorrhea by culture, and in our analyses, we additionally excluded those culture-positive for gonorrhea. We analyzed data from study sites that had complete clinical and laboratory data. Our study population was comprised of 577 nonpregnant asymptomatic females ages 17 to 25 seen at a gynecology clinic in San Francisco, California (UCSF) and a sexually transmitted diseases clinic in Brooklyn, New York (SUNY). Participants underwent chlamydial screening between October 1996 and October 1999 by the NAAT transcription-mediated amplification (TMA) (Amplified Chlamydia trachomatis Assay; Gen-Probe, San Diego, CA; a first-generation TMA, not APTIMA) and had pelvic examinations. Chlamydia positivity was defined as a positive TMA from at least one genital specimen type. Race/ethnicity data were unavailable; however, in general, the clinic population studied at UCSF was comprised mostly of Hispanics and at SUNY mostly of blacks.
Methods for specimen collection, transport, and chlamydial testing have been previously described.10 Genital specimens collected included first-catch urine, endocervical swabs, and self- and clinician-obtained vaginal swabs. Specimens were processed and chlamydial testing performed at an experienced laboratory at each study site, which also performed chlamydia culture on endocervical swabs using its standard procedure and TMA following the manufacturer's instructions. Institutional Review Board review and approval of the protocol was granted at both study sites as previously described.10
Analyses were conducted with Statistical Analysis Software, version 8.2 (SAS Institute, Cary, NC). Associations of pelvic examination findings with chlamydial TMA positivity or with chlamydial culture positivity in a subset of TMA-positive subjects were assessed by the Fisher exact test. Logistic regression analyses were used to estimate adjusted odds ratios (AOR) and 95% confidence intervals (CIs) controlling for potential confounding covariates.
Clinical Characteristics and Prevalence of Chlamydial Infection
The median age of the 577 asymptomatic females was 22 years (range, 17–25 years), and chlamydial TMA was positive in 68 (11.8%); the SUNY site had a higher C. trachomatis prevalence (55 of 389 [14%]) than the UCSF site (13 of 188 [7%]). Bacterial vaginosis was present in 94 (16.3%), trichomoniasis in 23 (4.0%), and vaginal candidiasis in 40 (6.9%) of the 577 females evaluated. Vaginal discharge (5.9%) was the most common pelvic examination finding followed by cervical ectopy, mucopurulent endocervical discharge, easily induced endocervical bleeding, cervical motion tenderness, and adnexal tenderness (all <5%) (Table 1). MPC signs of mucopurulent endocervical discharge and/or easily induced endocervical bleeding were present in 25 (4.3%) of women and differed by clinical site with 8 (2%) of SUNY patients versus 17 (9%) of UCSF patients having MPC.
Relationship of Pelvic Examination Findings With Chlamydial Transcription-Mediated Amplification Results
In univariate analyses, a positive chlamydial TMA result was associated with pelvic examination findings of vaginal discharge (11.8% vs. 5.1%; P = 0.048), mucopurulent endocervical discharge (7.4% vs. 1.6%; P = 0.012), and easily induced endocervical bleeding (7.4% vs. 1.6%; P = 0.012), but not cervical ectopy, cervical motion tenderness, or adnexal tenderness (Table 1). Mucopurulent endocervical discharge and/or easily induced endocervical bleeding was more significantly associated with a positive chlamydial TMA (13.2% vs. 3.1%; P = 0.001). After controlling for bacterial vaginosis, trichomoniasis, and candidiasis, a positive chlamydial TMA remains associated with mucopurulent endocervical discharge (AOR = 4.4; 95% CI = 1.3–14.3; P = 0.015), easily induced endocervical bleeding (AOR = 4.7; 95% CI = 1.5–15.4.; P = 0.010), and mucopurulent endocervical discharge and/or easily induced endocervical bleeding (AOR = 4.2; 95% CI = 1.7–10.2; P = 0.002), but not vaginal discharge (P >0.1).
Pelvic Examination Findings in Chlamydial Transcription-Mediated Amplification-Positive Females Stratified by Culture Results
Of the 68 chlamydial TMA-positive patients, 61 (89.7%) had a positive endocervical chlamydial culture. Although not a significant difference because of the small number of culture-negative subjects (N = 7), pelvic examination findings were absent in all 7 culture-negative subjects, whereas in culture-positive subjects, mucopurulent endocervical discharge and/or easily induced endocervical bleeding (14.8%; each 8.2% individually) and vaginal discharge (13.1%) occurred in approximately one per 7 patients (Table 2).
Abnormal pelvic examination findings were infrequent (<6%) in asymptomatic young women undergoing chlamydial screening by a chlamydial NAAT. However, the MPC signs mucopurulent endocervical discharge and/or easily induced bleeding, when present, predicted chlamydial infection, which is consistent with many other reported studies.2–7 In the 68 women with chlamydial infection, these MPC signs were present in 13%, which is much lower than reported in prior studies using chlamydial testing other than NAATs.7,8 A plausible explanation for this finding is that NAATs are more likely than non-NAAT (e.g., culture or DNA probe) to detect chlamydial infections with a small number of chlamydial organisms (i.e., inclusion forming units [IFU]) and because the quantity of chlamydial IFU correlate directly with MPC signs,11 NAAT will detect more chlamydial infections without MPC signs. This is further supported by our finding that MPC signs were only present in chlamydia-infected subjects who were chlamydia culture-positive but not chlamydia culture-negative, because the latter group likely had infections with a low number of chlamydial IFU; however, only a small proportion (7 of 68 [10%]) of chlamydia-infected women were culture-negative in this study, limiting confidence in this conclusion.
Our study findings have important implications in chlamydial screening practices.
There have been concerns that chlamydial screening methods using self-collection of urine or vaginal swabs and not a pelvic examination may miss a significant amount of morbidity and cannot be justified. Our findings refute that concern. We did not find a lot of disease that would only have been detected by a pelvic examination. In addition, it has been reported that pelvic examinations have negative attributes, including anxiety, pain, and less control over sample collection,12 which may be a barrier to chlamydial screening in females. Because only a small proportion (9 of 68) of these asymptomatic chlamydia-infected women had MPC signs and because PID signs were rare (2 of 68), chlamydial screening by NAAT on urine in higher-risk asymptomatic women, especially in nontraditional clinical settings where pelvic examinations are not always feasible (e.g., correctional facilities, schools, and so on), is certainly justified. Furthermore, of 25 subjects with MPC signs who may have been empirically treated for chlamydia (especially if there were risk factors for chlamydia or follow-up concerns), only 9 tested positive for chlamydia by TMA, and therefore up to 16 of the 25 subjects with MPC may have received unnecessary chlamydia treatment. Certainly, symptomatic women should have pelvic examinations, and in traditional clinical settings (e.g., sexually transmitted disease clinic, family planning, and so on) where pelvic examinations are more feasible, they should still be considered because detecting MPC or PID signs would allow earlier chlamydia treatment and the pelvic examination does provide an opportunity to detect other infections (e.g., genital herpes, genital warts, and early syphilis). Unfortunately, data on the presence of genital warts or testing for genital herpes or syphilis were not available; however, 2 patients seen at SUNY and 3 at UCSF had ulcers noted on examination.
The retrospective design of this study is a limitation in that some demographic and clinical information, which may influence the frequency of MPC findings or risk for chlamydial infection, was not available. It has been reported that chlamydial IFU counts may be influenced by race,13 which, as discussed previously, could influence the presence of MPC signs; however, race, IFU, or semiquantitative chlamydia culture data were not available to incorporate in our analyses. We also did not have data on hormonal contraception use, in which prior studies have been contradictory on whether oral contraception is a risk factor for chlamydia, but evidence is building that depot-medroxyprogesterone is associated with increased risk of chlamydial infection.14–16 Although we knew the proportion of subjects with ectopy in our study, we did not have more information on whether some may have had hypertropic ectopy (edematous, congested, and bleeds easily), which also has been reported to predict chlamydial infection7,8 and is considered by some to be a sign of MPC. Finally, we did not have information on prior chlamydial infection, and it is possible subjects with prior chlamydia may develop some degree of protective immunity and have less inflammation with repeat infection and hence may be less likely to have MPC signs.
In conclusion, pelvic examination findings of MPC were infrequent and of PID were rare in young asymptomatic women screened by chlamydial NAAT. This should increase confidence for screening in nontraditional settings where pelvic examinations are not feasible and reduce the concern that significant clinical disease is being missed by not performing pelvic examinations. Screening in such settings will be an important component of our future efforts toward the control of chlamydial infections.
1. Stamm WE. Chlamydia trachomatis
—The persistent pathogen: Thomas Parran Award Lecture. Sex Transm Dis 2001; 28:684–689.
2. Brunham RC, Paavonen J, Stevens CE, et al. Mucopurulent cervicitis—The ignored counterpart in women of urethritis in men. N Engl J Med 1984; 311:1–6.
3. Lindner LE, Geerling S, Nettum JA, et al. Clinical characteristics of women with chlamydial cervicitis. J Reprod Med 1988; 33:684–690.
4. Critchlow CW, Wolner-Hanssen P, Eschenbach DA, et al. Determinants of cervical ectopia and of cervicitis: Age, oral contraception, specific cervical infection, smoking, and douching. Am J Obstet Gynecol 1995; 173:534–543.
5. Paavonen J, Critchlow CW, DeRouen T, et al. Etiology of cervical inflammation. Am J Obstet Gynecol 1986; 154:556–564.
6. Paavonen J, Vesterinen E. Chlamydia trachomatis
in cervicitis and urethritis in women. Scand J Infect Dis Suppl 1982; 32:45–54.
7. Paavonen J, Stevens CE, Wolner-Hanssen P, et al. Colposcopic manifestations of cervical and vaginal infections. Obstet Gynecol Surv 1988; 43:373–381.
8. Harrison HR, Costin M, Meder JB, et al. Cervical Chlamydia trachomatis
infection in university women: Relationship to history, contraception, ectopy, and cervicitis. Am J Obstet Gynecol 1985; 153:244–251.
9. Marrazzo JM, Johnson RE, Green TA, et al. Impact of patient characteristics on performance of nucleic acid amplification tests and DNA probe for Chlamydia trachomatis
genital infection in women. J Clin Microbiol 2005; 43:577–584.
10. Schachter J, McCormack WM, Chernesky MA, et al. Vaginal swabs are appropriate specimens for diagnosis of genital tract infection with Chlamydia trachomatis.
J Clin Microbiol 2003; 41:3784–3789.
11. Geisler WM, Suchland RJ, Whittington WLH, Stamm WE. Quantitative culture of Chlamydia trachomatis
: Correlation of inclusion-forming units produced in culture with clinical manifestations of urogenital disease. J Infect Dis 2001; 184:1350–1354.
12. Serlin M, Shafer M, Tebb K, et al. What sexually transmitted screening method does the adolescent prefer? Arch Pediatr Adolesc Med 2002; 156:588–591.
13. Eckert LO, Suchland RJ, Hawes SE, Stamm WE. Quantitative Chlamydia trachomatis
cultures: Correlation of chlamydial inclusion forming units with serovar, age, gender, and race. J Infect Dis 2000; 182:540–544.
14. Morrison CS, Bright P, Wong EL, et al. Hormonal contraceptive use, cervical ectopy, and the acquisition of cervical infections. Sex Transm Dis 2004; 31:561–567.
15. Jacobson DL, Peralta L, Farmer M, et al. Relationship of hormonal contraception and cervical ectopy as measured by computerized planimetry to chlamydial infection in adolescents. Sex Transm Dis 2000; 27:313–319.
16. Dayan L, Donovan B. Chlamydia, gonorrhoea, and injectable progesterone. Lancet 2004; 364:1387–1388.