INFECTIONS WITH Neisseria gonorrhoeae, Chlamydia trachomatis, and Trichomonas vaginalis are common in women living in central and northern Australia. 1,2 Complications due to N gonorrhoeae and C trachomatis such as pelvic inflammatory disease and infertility are well recognized, 3,4 and there is increasing evidence that infection with T vaginalis can result in adverse pregnancy outcome. 5 Furthermore, infection with all three pathogens has been associated with increased transmission of HIV. 6,7
Making a definitive diagnosis of N gonorrhoeae, C trachomatis and T vaginalis infection in women living in remote areas of central Australia is hampered by several obstacles. There can be long delays before specimens reach the regional laboratory, resulting in deterioration of the specimen quality and a reduction in the sensitivity of the test. Practitioners working in remote clinics may not be experienced in carrying out pelvic examinations, and as a result inadequate specimens may be obtained. Women may also be reluctant to access the clinic or undergo a pelvic examination for a variety of reasons, such as shame and the fear of loss of privacy and confidentiality in a small community.
The use of more sensitive and rapid diagnostic tests such as polymerase chain reaction (PCR) and less invasively collected specimens such as urine has been shown to increase the detection of N gonorrhoeae and C trachomatis among men and women in central Australia. 8 Currently, regional protocols for women in central Australia recommend practitioner-collected endocervical swabs as the best specimens for PCR detection of N gonorrhoeae and C trachomatis. However, urine samples for PCR testing are being used increasingly in remote areas of the region for screening, for opportunistic testing, and for women who refuse an examination. While urine is both an acceptable and easy specimen to collect, a recent study conducted in Northern Australia found urine tests missed a significant proportion of N gonorrhoeae compared to tampon PCR, although for Chlamydia and Trichomonas detection, urine and tampon samples were equally sensitive. 9 In a further study in women presenting for legal termination of pregnancy in Melbourne in a population with relatively low prevalence of chlamydial infection (chlamydia; 2.8%), patient-collected samples (tampon and first-void urine [FVU]) were found to be as reliable as physician-collected endocervical swabs for PCR detection. 10 In other studies, self-collected specimens such as tampons and vaginal swabs have also been found to be acceptable and sensitive in testing for sexually transmitted infections (STIs) in women. 1,9,11,12
Infection with T vaginalis is also common in women in central and northern Australia. 1,13 While PCR tests for detection of T vaginalis have been developed, they are not commercially available and to date have been used in central Australia for research purposes only. Currently, T vaginalis is detected by microscopy of wet preparations of high vaginal swabs. This method of detection is significantly less sensitive than PCR and is also affected by suboptimal transport conditions and delays that occur in testing in remote communities. 14
In order to determine the best specimen for detection of N gonorrhoeae and C trachomatis by PCR and hence advise about regional protocols, we designed a study to compare PCR testing of four different specimens: urine, self-collected vaginal swab, self-collected tampon, and a practitioner-collected endocervical swab specimens. In addition, for T vaginalis detection by PCR, we compared two samples—self-collected vaginal swab and tampon—as well as a practitioner-collected swab for microscopy and culture.
This study was conducted over a 16-month period between February 1998 and May 1999. It was a collaboration between health services in remote areas of Australia, including the Alice Springs Sexual Health Unit (Territory Health Services), Nganampa Health Council, Ngaanyatjarra Health Service, and Alice Springs Remote Health Service. Ethics approval was granted by the Alice Springs Institutional Ethics Committee. A Memorandum of Understanding was signed by the investigators of the study and the participating health services and laboratories regarding collection and use of the data. All women attending the local clinic who required a pelvic examination as part of a symptomatic or asymptomatic check-up were invited to participate. Participation was voluntary, and patients gave written or verbal consent before examination. Information in the local language about the study and PCR testing was given by female Aboriginal health workers. All women who participated in the study as well as those who declined were cared for in the same way, according to regional protocols.
Specimen Collection and Transport
A total of six specimens, including three self-collected samples, were collected from each participant. Participants were given instructions on appropriate collection of specimens before the speculum examination. Women were asked to pass 10 to 20 ml of FVU into a sterile jar and then to insert 2 to 4 cm into the low vagina a swab, which was circulated for several seconds and then removed and placed into a dry plastic tube. This was followed by collection of a tampon specimen (Meds Regular Tampons; Johnson & Johnson™), which was inserted into the vagina, immediately removed, and placed into a container of transport medium, as previously described. 1 Then the women were reviewed by a female practitioner, a speculum examination was performed, and one high vaginal and two endocervical swabs were collected. The high vaginal swab and one of the endocervical swabs were immediately rolled onto a glass slide at the bedside and subsequently placed into Stuart's transport medium. The second endocervical swab (for PCR testing) was inserted into a sterile plastic tube containing no transport medium.
All specimens were dispatched to the regional laboratory (Western Diagnostic Pathology, Alice Springs), where conventional tests for microscopy and culture were performed on high vaginal swabs for T vaginalis and on endocervical swabs for N gonorrhoeae. The second endocervical swab, the self-collected vaginal swab, urine, and tampon specimens were transported to the Molecular Microbiology Laboratory of the Royal Women's Hospital in Melbourne, Australia, for nucleic acid amplification testing (NAAT). All these specimens for PCR testing were sent in a polystyrene foam container with an ice block to the regional laboratory and subsequently were transported to the laboratory in Melbourne. Samples were kept at refrigeration temperatures in the laboratory before extraction.
In the laboratory a wet preparation was prepared from the high vaginal swabs in transport media for examination by microscopy for the presence of T vaginalis flagellates and then inoculated into trichomonas broth (Oxoid, Hampshire, UK) and examined daily subsequent to incubation at 35 °C for 48 hours. The endocervical swabs in transport media were subsequently plated onto gonococcal selective agar (Oxoid) and gonococcal nonselective agar (Oxoid) and incubated at 35 °C for 48 hours in a 5% CO2 incubator. Identification of N gonorrhoeae was made by means of the oxidase test, Phadebact monoclonal gonococcal test (Boule Diagnostics, Sweden), and Gonocheck–II (EY Laboratories, San Mateo, CA).
Sample Preparation for NAAT
Urine, self-collected vaginal swabs, and practitioner-collected endocervical swabs were processed according to the protocols for Roche COBAS Amplicor (Roche Diagnostics, Branchburg, NJ).
For tampon samples, after collection of cell pellets, DNA was extracted from a 20-μl aliquot with a QIAamp DNA Purification Kit (Qiagen, Valencia, CA), per the manufacturer's instructions. DNA was eluted in 200 μl of elution buffer (Qiagen). Processing of DNA was done by mixing 20 μl of DNA, 5 μl of COBAS Lysis Solution, and 25 μl of specimen diluent, followed by a 10-minute room temperature incubation before NAAT. 15 Subsequent to processing of the self-administered swab by COBAS protocols, 200 μl of the processed sample was extracted for DNA by the Qiagen method as described above. The extracted DNA from this sample was used for T vaginalis PCR.
A 50-μl aliquot of processed samples, including urine, self-collected vaginal swabs, tampons, and practitioner-collected endocervical swabs, was amplified for C trachomatis, N gonorrhoeae and an internal control by Roche Cobas Amplicor. Internal control–negative results were confirmed by repeated testing. When only one sample from a participant was positive for C trachomatis, a confirmatory test was performed with a second PCR with primers directed to a major OMP. 16 Participants were considered positive for C trachomatis if positive results were obtained with samples from at least two sites or with confirmation by OMP-PCR when only one sample was positive.
All PCR-positive N gonorrhoeae specimens, whether culture-positive or -negative, were confirmed as such with a Roche 16S confirmatory assay 17 (Roche Diagnostics). Participants were considered positive for N gonorrhoeae if any one sample was confirmed positive by 16S PCR.
Amplification for T vaginalis and β-globin DNA sequences was performed in noncommercial assays and utilized extracted DNA from tampon and self-collected swab samples. Each PCR reaction consisted of a 20-μl aliquot of extracted DNA, 1× reaction buffer (50 mmol/l KCl, 10 mmol/l Tris [pH, 8.3]), 200 μmol/l each of dATP, dGTP, and dCTP, and 190 μmol/l of dTTP, 10 μmol/l of digoxigenin-dUTP, and 2.5 units of the heat-stable Amplitaq Gold Polymerase (Perkin Elmer Cetus, NJ) in a total of 50 μl. Amplification reaction for T vaginalis DNA sequences consisted of 50 ρmol of each primer, TVA5–TVA6, 18 and 1.5 mmol/l of MgCl2. Each reaction was amplified 35 cycles with parameters of 94 °C for 1 minute, 47 °C for 1 minute, and 67 °C for 1 minute. Amplification of β-globin sequences consisted of the addition of 5 ρmol of each of the primers GH20–PC04 18 and 4 mmol/l of MgCl2. β-globin amplification is included as a positive internal control; it is directed to amplify a human β-globin product of 260 bp. Each reaction was amplified 35 cycles with parameters of 95 °C for 1 minute, 55 °C for 1 minute, and 72 °C for 1 minute. Positive clinical specimens by culture for T vaginalis were used as positive controls. Strict procedures avoiding specimen contamination and carryover were followed.
PCR products were hybridized and detected by Enzymun-Test DNA Detection Assay (Boehringer Mannheim) with use of the automated ES 300 analyzer, according to the manufacturer's protocol. Biotin-labeled oligonucleotide capture probes TB (5′GACCTCTAGAAGAAGACTCAG3′) 1 and PC03 19 were used for detection of T vaginalis and β-globin sequences, respectively. When T vaginalis was detected in only one specimen by PCR, samples were tested by amplification with a second primer pair and subsequent hybridization, as described by Kengne et al. 15 The primers used amplify a 450-bp region from a 2000-bp repeated DNA fragment of T vaginalis. Participants were considered positive for T vaginalis if a positive result occurred by any of the following methods: wet preparation, culture, PCR positive for both swab and tampon, and PCR positive for either swab or tampon and confirmed by the second primer pair.
Sensitivity and specificity for each specimen type tested were compared to the final result per participant, as obtained subsequent to confirmation of results for each microorganism and as defined above. For C trachomatis and N gonorrhoeae, observations were excluded from the analysis if only one of the four specimens tested by NAAT was assessable, as determined by presence of a positive Roche internal control. In the case of T vaginalis, β-globin results were used to determine assessability for tampon and swab specimens. Confidence intervals were calculated on the basis of binomial distributions. 20 Comparison of sensitivity and specificity was done with the McNemar test. 20
A total of 318 (262 Aboriginal and 56 non-Aboriginal) women participated in the study. Seventy-four of these women were tested at an urban sexual health clinic, and 244 were tested at one of several remote clinics in central Australia. Five women where excluded from the analysis, as only one specimen was obtained. Overall, a tampon specimen was collected from 306, self-collected vaginal swabs from 307, urine from 292, and endocervical and high vaginal swabs from 302 women. Nonassessability, as determined by a negative Roche internal control (a marker of inhibition for PCR in the sample), was observed for 21 urine, 1 tampon DNA, 32 self-collected vaginal swab, and 57 practitioner-collected endocervical swab specimens.
The overall prevalence of STIs among the participants was as follows:C trachomatis, 11.5%;N gonorrhoeae, 11.8%; and T vaginalis, 24.6%. Detection of C trachomatis was most sensitive with tampon specimens (100%) and practitioner-collected endocervical swabs (92.3%) and was least sensitive with urine samples (72.7%;P = 0.01) (Table 1). Overall, 26 patients had more than one infection.
N gonorrhoeae was detected with the highest sensitivity in PCR of self-collected tampon specimens (97.2%), followed by practitioner-collected endocervical swabs (92.6%), self-collected vaginal swabs (71.9%), and urine (31.2%) (Table 2). The sensitivity of urine for detection of N gonorrhoeae by PCR was significantly lower than that of PCR of tampon (P < 0.0001), endocervical swab (P < 0.001), and self-collected swab specimens (P = 0.01). A statistically significant difference was also found when comparing self-collected swab to tampon specimens (P = 0.01).
While the aim of this study was primarily to compare 4 different clinical samples (3 patient-collected and 1 practitioner-collected) for PCR detection of N gonorrhoeae and C trachomatis, 2 samples (both patient-collected) were also evaluated for PCR detection of T vaginalis. Tampon specimens and self-collected vaginal swabs were tested using an in-house PCR test. Nineteen specimen sets were not analyzed, as one of the specimens was determined nonassessable by a negative β-globin result. The sensitivity and specificity of self-collected vaginal swabs versus tampons for detection of T vaginalis were 87.7% and 100%, respectively (Table 3).
While microscopy and culture for T vaginalis were also carried out, specimens were subjected to suboptimal transport conditions and long transport delays, which would have affected sensitivity and as a result cannot be regarded as a true “gold standard.” Inoculation of high vaginal swabs into trichomonas culture media at the laboratory did not improve detection of T vaginalis above that detected by microscopy. Sensitivity of microscopy and culture versus PCR for detection of T vaginalis was 52.4% (Table 4). In the comparison of conventional tests (microscopy and culture) with PCR for detection of N gonorrhoeae, the sensitivity was 53.6% (Table 4). No samples positive by culture were negative by PCR.
This study compared three self-collected specimens and one practitioner-collected specimen for the detection of STIs by PCR. All specimens were tested by the same commercially available PCR method for detection of C trachomatis and N gonorrhoeae. The self-collected vaginal swab and tampon specimens were also tested by an in-house PCR assay for detection of T vaginalis. The high prevalence of STIs (C trachomatis, 11.5%;N gonorrhoeae, 11.8%; and T vaginalis, 24.6%) detected among women in this study is similar to findings of previous studies conducted in central and northern Australia. 1,13
Overall, 37 patients had one or more samples not collected; there were only nine positive findings for various microorganisms for these patients. Exclusion of these patients from the analysis did not affect the overall results (data not shown).
Tampon specimens were tested both before and after extraction of DNA (data not shown). As has been reported following our previous studies, 21 the addition of DNA extraction from the clinical sample before amplification significantly increases the sensitivity for the detection of both organisms by PCR. Much of this can be explained by the fact that there are fewer nonassessable samples than when DNA is not extracted from crude preparations, which suggests the presence of inhibitors in crude tampon pellets. As shown in Table 1, samples that were nonassessable from urine totalled 21; from self-collected vaginal swabs, 30; and from practitioner-collected endocervical swabs, 49. Nonassessable tampon samples (DNA extracted) numbered only one. Tampons tested without DNA extraction (data not shown) resulted in 36 nonassessable specimens. The sensitivity of tampon specimens before DNA extraction for C trachomatis was 96.7%, and for N gonorrhoeae, 84.8%. We believe this additional step of DNA extraction reduces opportunities for inhibition of the PCR reaction, and it is standard practice in our laboratory. 21
This study also demonstrated that urine specimens transported from remote settings are not as sensitive as other sampling methods for detection of C trachomatis and N gonorrhoeae by PCR. The lengthy and suboptimal transport conditions, as well as inhibitors in urine specimens, could have contributed to the low sensitivity of urine for detection of N gonorrhoeae in this study.
Among men, various studies have reported a sensitivity range for PCR detection of N gonorrhoeae in urine versus urethral swab specimens between 83.8% and 100%. 22–24 However, for detection of N gonorrhoeae in women, only a few studies have compared the sensitivity of PCR of urine and swabs. A study conducted in two urban STD clinics in which specimens were stored at 4 °C and processed within 96 hours showed a sensitivity of 90% for N gonorrhoeae detection by PCR in urine versus endocervical swab specimens. 23 In contrast, a study conducted in remote communities in northern Australia showed a sensitivity of 42% for N gonorrhoeae detection by PCR in urine specimens versus tampons. 1 Average delay between collection and processing of specimens in that study was 7 days (range, 1–26) for rural specimens and 9 days (range, 2–31) for urban specimens, 1 whereas in this study, longer transport delays were encountered. An average of 13 days (range, 3–49) passed from the time of collection to processing, which may have contributed to the low sensitivity for detection of N gonorrhoeae in urine samples (31.2%).
The quality of urine specimens may also be affected by inhibitors present in the urine, which can cause a reduction in sensitivity. 25,26 Refrigerating or freezing urine specimens can improve sensitivity by reducing inhibitors. 25 While urine specimens were to be transported in polystyrene foam containers with ice bricks and refrigerated on arrival at the regional laboratory, it was noted that none of the specimens were frozen. Furthermore, it was not known what proportion of urine specimens were refrigerated before transportation to the laboratory.
In order to see whether improved storage and transportation conditions for urine specimens may improve sensitivity, participating health services performed a pilot study to collect more data after completion of the study. Between April and May 2000, women living in participating remote communities were asked to collect a FVU sample and then a self-collected vaginal swab specimen when presenting to their health clinic for an annual STI check-up. Urine specimens were frozen within 2 hours of collection and processed within 7 days of collection. Among the 366 sets of specimens collected, the sensitivity of PCR for detection of N gonorrhoeae and C trachomatis in urine versus self-collected swab specimens was 77.8% and 84%, respectively, a marked improvement over the sensitivity for detection of N gonorrhoeae in the original study.
In this study in a remote setting, PCR testing of self-collected swab and tampon specimens was superior to conventional microscopy and culture for the detection of N gonorrhoeae and T vaginalis. While suboptimal transport conditions and delays could account for the low sensitivity of culture for N gonorrhoeae, it is necessary to attempt to culture N gonorrhoeae at least periodically (or on a patient-recall basis when treatment for a PCR-positive result is needed) to monitor trends in antibiotic sensitivity in the region and advise on empirical treatment.
PCR for detection of T vaginalis is currently not commercially available, but given the high prevalence of such infection and low sensitivity of microscopy, it is important for PCR to become more widely available for use in remote areas such as those in Australia.
While detection of C trachomatis and N gonorrhoeae in self-collected swabs was less sensitive than in tampons or endocervical swabs, this difference was only statistically significant when self-collected swabs were compared with tampons for detection of N gonorrhoeae (P = 0.01).
Like tampons, self-collected swabs provide women with an alternative way of testing for STI that is both acceptable and sensitive. Self-collected swabs have the advantage over tampons of not requiring transport medium and being less time-consuming to process at the laboratory. While self-collected swabs and tampons do not replace a speculum examination and Papanicolaou smear, there are many circumstances in which such sampling methods are an acceptable and sensitive alternative to doing an examination. Alternative methods of testing for STI may be needed when a woman refuses an examination, for women in late pregnancy, and if the practitioner is inexperienced in performing a speculum examination. Self-collection of specimens may be particularly useful for young women who may not yet need a Papanicolaou smear examination and may be discouraged from attending the clinic by the prospect of a speculum examination yet are at risk of acquiring an STI.
1. Tabrizi SN, Paterson B, Fairley CK, Bowden FJ, Garland SM. A self-administered technique for the detection of sexually transmitted diseases in remote communities. J Infect Dis 1997; 176: 289–292.
2. Miller PJ, Torzillo PJ, Hateley W. Impact of improved diagnosis and treatment on prevalence of gonorrhoea and chlamydial infection in remote Aboriginal communities on Anangu Pitjantjatjara Lands. Med J Aust 1999; 170: 429–432.
3. Westrom L, Eschenbach D. Pelvic inflammatory disease. In: Holmes KK, Sparling PF, Mardh PA, eds. Sexually Transmitted Diseases. New York: McGraw-Hill, 1999: 783–809.
4. Cates W Jr, Brunham RC. Sexually transmitted disease and infertility. In: Holmes KK, Sparling PF, Mardh PA, eds. Sexually Transmitted Diseases. New York: McGraw-Hill, 1999: 1079–1087.
5. McGregor JA, French JI, Parker R, et al. Prevention of premature birth by screening and treatment for common genital tract infections: results of a prospective controlled evaluation. Am J Obstet Gynecol 1995; 173: 157–167.
6. Laga M, Manoka A, Kivuvu M, et al. Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. AIDS 1993; 7: 95–102.
7. Gregson S, Mason PR, Garnett GP, et al. A rural HIV epidemic in Zimbabwe? Findings from a population-based survey. Int J STD AIDS 2001; 12: 189–196.
8. Skov SJ, Miller PJ, Hateley W, Bastian IB, Davis J, Tait PW. Urinary diagnosis of gonorrhoea and chlamydia in men in remote Aboriginal communities. Med J Aust 1997; 166: 468–471.
9. Tabrizi SN, Paterson BA, Fairley CK, Bowden FJ, Garland SM. Comparison of tampon and urine as self-administered methods of specimen collection in the detection of Chlamydia trachomatis, Neisseria gonorrhoeae
and Trichomonas vaginalis
in women. Int J STD AIDS 1998; 9: 347–349.
10. Garland SM, Tabrizi S, Hallo J, Chen S. Assessment of Chlamydia trachomatis
prevalence by PCR and LCR in women presenting for termination of pregnancy. Sex Transm Infect 2000; 76: 173–176.
11. Tabrizi S, Chen MS, Borg AJ, et al. Patient administered tampon-collected genital cells in the assessment of Chlamydia trachomatis
infection using polymerase chain reaction. Sex Transm Dis 1996; 23: 494–497.
12. Witkin SS, Inglis SR, Polaneczky M. Detection of Chlamydia trachomatis
and Trichomonas vaginali
s by polymerase chain reaction in introital specimens from pregnant women. Am J Obstet Gynecol 1996; 175: 165–167.
13. Voolman T, Morey F, Rich G. Trichomoniasis is a problem for Aboriginal women: fact or fiction? Venereology 1995; 8: 34–36.
14. Paterson BA, Tabrizi SN, Garland SM, Fairley CK, Bowden FJ. The tampon test for trichomoniasis: a comparison between conventional methods and a polymerase chain reaction for Trichomonas vaginalis
in women. Sex Transm Infect 1998; 74: 136–139.
15. Kengne P, Veas F, Vidal N, Rey JL, Cuny G. Trichomonas vaginalis
: repeated DNA target for highly sensitive and specific polymerase chain reaction diagnosis. Cell Mol Biol (Noisy-le-grand) 1994; 40: 819–831.
16. Holland SM, Gaydos CA, Quinn TC. Detection and differentiation of Chlamydia trachomatis, Chlamydia psittaci
and Chlamydia pneumoniae
by DNA amplification. J Infect Dis 1990; 162: 984–987.
17. Farrell DJ. Evaluation of AMPLICOR Neisseria gonorrhoeae
PCR using cppB nested PCR and 16S rRNA PCR. J Clin Microbiol 1999; 37: 386–390.
18. Riley DE, Roberts MC, Takayama T, Krieger JN. Development of a polymerase chain reaction-based diagnosis of Trichomonas vaginalis.
J Clin Microbiol 1992; 30: 465–472.
19. Resnick RM, Cornelissen MT, Wright DK, et al. Detection and typing of human papillomavirus in archival cervical cancer specimens by DNA amplification with consensus primers. J Natl Cancer Inst 1990; 82: 1477–1484.
20. Altman DA. Practical Statistics for Medical Research. London: Chapman and Hall, 1991.
21. Tabrizi SN, Fairley CK, Chen S, et al. Evaluation of patient-administered tampon specimens for Chlamydia trachomatis
and Neisseria gonorrhoeae.
Sex Transm Dis 2000; 27: 133–137.
22. Higgins SP, Klapper PE, Struthers JK, et al. Detection of male genital infection with Chlamydia trachomatis
and Neisseria gonorrhoea
using an automated multiplex PCR system (Cobas Amplicor). Int J STD AIDS 1998; 9: 21–24.
23. Crotchfelt KA, Welsh LE, DeBonville D, Rosenstraus M, Quinn TC. Detection of Neisseria gonorrhoe
a and Chlamydia trachomatis
in genitourinary specimens from men and women by a coamplification PCR assay. J Clin Microbiol 1997; 35: 1536–1540.
24. Palladino S, Pearman JW, Kay ID, et al. Diagnosis of Chlamydia trachomatis
and Neisseria gonorrhoea
genitourinary infections in males by the Amplicor PCR assay of urine. Diagn Microbiol Infect Dis 1999; 33: 141–146.
25. Rosenstraus M, Wang Z, Chang SY, DeBonville D, Spadoro JP. An internal control for routine diagnostic PCR: design, properties, and effect on clinical performance. J Clin Microbiol 1998; 36: 191–197.
26. Mahony J, Chong D, Jang D, et al. Urine specimens from pregnant and nonpregnant women inhibitory to amplification of Chlamydia trachomatis
nucleic acid by PCR, ligase chain reaction, and transcription-mediated amplification: identification of urinary substances associated with inhibition and removal of inhibitory activity. J Clin Microbiol 1998; 36: 3122–3126.