UROGENITAL CHLAMYDIA TRACHOMATIS INFECTIONS are widespread worldwide,1 and many symptomatic and asymptomatic individuals are tested annually.2 Testing is performed to diagnose and treat symptoms like discharge and dysuria. Other important objectives are to prevent pelvic inflammatory disease (PID)3 and long-term complications like ectopic pregnancies and female infertility.4
The association between C. trachomatis infections and ectopic pregnancy and infertility has been supported in several studies with different designs. Thus, C. trachomatis has been shown to be 1 of the main causative agents in symptomatic PID.5 The role of symptomatic PID in infertility and ectopic pregnancy was ascertained in a Swedish long-term follow-up study,4 which showed a higher rate of ectopic pregnancy and infertility among women with (laparoscopically verified) PID than among women having no PID. This study, however, was performed before C. trachomatis testing was introduced and therefore does not specifically address the role of C. trachomatis infections. Later studies of symptomatic PID caused by C. trachomatis have indicated that prompt antibiotic treatment may largely prevent long-term complications in the form of infertility.5–7 Several seroepidemiologic cross-sectional studies8–14 support an association between asymptomatic untreated C. trachomatis infection and fertility problems, and because this association is also found among women without a history of symptomatic PID, it presumably reflects that subclinical PID causes long-term complications.15 It should be noted, however, that the serologic tests used for the case–control studies are problematic regarding both sensitivity and specificity.16
It has recently been documented that C. trachomatis testing generates anxiety about the reproductive consequences among women receiving a positive test result.17 This stresses the need to assess the long-term complications among those diagnosed with an infection to be able to provide relevant information and advice. Earlier studies have not been undertaken to ascertain the incidence rates of ectopic pregnancy or reduced fertility in women having a C. trachomatis-positive test result, possibly because of the often long timespan between the infection and the complications. To equip clinicians taking care of C. trachomatis-infected women with a better prognostic tool, we analyzed ectopic pregnancies and births in a cohort of women previously tested for C. trachomatis.
Materials and Methods
We conducted a historical follow-up study among women tested for C. trachomatis from 1984 to 1993 in the County of Northern Jutland (Denmark) with a population of approximately 500,000, corresponding to 9% of the total Danish population. Tax-supported health care is provided by the National Health Service to all citizens free of charge. Linkage of Danish health registers is simple and enjoys high validity. Linkage is achieved through the use of the 10-digit civil registry number (CPR number), which encodes gender and date of birth and is unique to every Danish citizen.
The study population consisted of 22,264 tested women from whom a total of 31,175 samples had been obtained. The samples were analyzed at the Department of Clinical Microbiology, Aalborg Hospital. Laboratory records were kept in paper files and these data were entered into a database in 2002.
Women in the study population were offered testing in general practice in a sexually transmitted diseases (STD) clinic or in a hospital. Testing was recommended on 4 indications: before transcervical procedure (abortion, insertion of an intrauterine device, or a hysterosalpingography), as a result of the presence of symptoms, and as part of a partner notification scheme or an opportunistic screening program.2 The reasons for obtaining a test were not recorded. More than 85% of the samples in Denmark are obtained in general practice. Samples were obtained only from the cervix, from the urethra only, or from both sites.
Chlamydia trachomatis Tests
The Department of Clinical Microbiology, Aalborg Hospital has been providing C. trachomatis tests to the county’s general practitioners and hospitals since the early 1980s. The tests measured direct immunofluorescence (Microtrak EIA; Syva) and enzyme-linked immunosorbent assay (ELISA) (Chlamydiazyme; Abbott). With the exception of a brief period in 1988, all positive or “gray-zone” ELISA results were confirmed either by direct immunofluorescent staining of a pellet prepared from the sample or by addition of a blocking antibody to the Chlamydiazyme test.
The outcomes evaluated in our study were ectopic pregnancies and births (alive or stillborn). We also identified the diagnoses of spontaneous abortion and induced abortion for use in the statistical analyses (described subsequently). All the diagnoses indicating pregnancy (ectopic pregnancy, spontaneous abortion, induced abortion, and births) were identified in the Northern Jutland County Hospital Discharge Registry. Data from this register are transferred to the national Danish Hospital Discharge Registry, which comprises data on 99.4% of all discharges from Danish hospitals. Data in this register include CPR numbers, dates of admission and discharge, the surgical procedures performed, and relevant diagnoses classified according to the Danish versions of International Classification of Diseases, 8th Revision (ICD-8) until the end of 1993 and according to the ICD-10 since 1994. The registry has been evaluated and it was found that 87% of codes for diagnoses were in accordance with the hospitals case records for pediatrics, general surgery, and gynecology/obstetrics.18 In our study, the codes used for births were: ICD-8: 650–666; ICD-10: 080, 081, 082, 083, 084; codes used for ectopic pregnancy were: ICD-8: 631; ICD-10: 000; codes used for spontaneous abortions were: ICD-8: 643.8, 643.9, 634.61, 645.1; ICD-10: 002, 003; and codes used for induced abortions were: ICD-8: 640, 641,642; ICD-10: 004, 005, 006.
We chose to consider only first pregnancies among women in the study population to analyze independent data. A first pregnancy was defined as a pregnancy diagnosis (ectopic pregnancy, induced abortion, spontaneous abortion, or birth) registered as a first-time pregnancy in the Northern Jutland County Hospital Discharge Registry. Women having had a pregnancy before their first C. trachomatis test were not at risk for a first pregnancy after the test and were therefore excluded from further analysis. Women not tested before the age of 43 were also excluded because we considered their risk for a pregnancy to be low.
To be sure only to include women actually at risk for ectopic pregnancy, we chose only to include those registered with a pregnancy during the follow-up (until January 1, 2003). Thus, women not registered with a pregnancy diagnosis during follow-up were excluded.
We used Cox regression with time-dependent covariates to estimate the hazard ratio for an ectopic pregnancy. Women were included at the date of their first C. trachomatis test. If this test was positive, the woman remained in the group with positive test results, even if she later had a negative test. If a woman first had a negative test and later a positive test, she changed exposure group at the date of the positive test result. Follow-up ended at the date of hospitalization for an ectopic pregnancy, at the date of giving birth to a child, or at the date of a spontaneous or an induced abortion, whichever came first. The number of ectopic pregnancies per 100 person-years at risk was calculated both for the group of C. trachomatis-positives and for the group of C. trachomatis-negatives, and the corresponding incidence rates were calculated. To adjust for maternal age, we included age as a time-dependent covariate. Finally, crude and adjusted hazard ratios with 95% confidence intervals (CIs) were calculated.
We analyzed the time to birth as a measure of the reproductive capacity assuming that infertile women waited longer before giving birth to a child than women with normal fertility. All women having a C. trachomatis test before their first pregnancy and before the age of 43 were included in the cohort. Thus, we also included women who did not become pregnant during follow-up.
As for the analysis of ectopic pregnancies, we used Cox regression with time-dependent covariates, and women were included at the date of their first C. trachomatis test and changed exposure status over time as for the analysis of ectopic pregnancies. Follow-up ended at the date of hospitalization for an ectopic pregnancy, at the date of giving birth to a child or having a spontaneous or an induced abortion, at the date of moving out of North Jutland County, at the age 43 or on January 1, 2003, whichever came first. The number of births per 100 person-years at risk and the corresponding incidence risks and hazard ratios were calculated the same way as for the ectopic pregnancies.
This study was approved by The Danish Data Protection Agency.
Among the 22,264 women in the study population, we excluded 6945 women having their first pregnancy before registration of their first C. trachomatis test and another 1626 women because they were 43 years or older before their first C. trachomatis test (Fig. 1). The remaining 13,693 women (in the following, referred to as the study cohort) had a total of 18,452 C. trachomatis tests performed. Samples were simultaneously obtained from the cervix and urethra in 1715, from the cervix only in 16,619, and from the urethra only in 118. In the study cohort, we identified 1882 infected women: 1763 were infected only once, 111 were infected twice, 7 were infected 3 times, and 1 woman 4 times. The infection prevalence in samples from the study cohort was 10.9% (2010 of 18,452). Among women in the study cohort, 13.7% (1882 of 13,693) experienced at least 1 C. trachomatis-positive test result. Because of the low number of women having more than 1 infection registered in the database, we were not able to analyze the impact of reinfections on ectopic pregnancies and time to birth.
The total number of outcome observations among the women included for further analysis (n = 13,693) was 175 ectopic pregnancies and 4067 births (Table 1). The mean maternal age was 29.7 years for those who experienced an ectopic pregnancy as their first pregnancy and 27.7 years for those who gave birth to a child as a consequence of their first pregnancy (data not shown).
The analysis of ectopic pregnancies included 906 C. trachomatis-positive and 5166 C. trachomatis-negative women. Among the positives, 14 (1.5%; 95% CI, 0.7–2.4%) experienced an ectopic pregnancy during follow up and the corresponding number among the negatives was 161 (3.1%; 95% CI, 2.64–3.59%) (Table 1). The mean observation time before pregnancy was 3.7 years (data not shown). The incidence rate of ectopic pregnancy among the positives was 0.35 per 100 person-years (14 of 3950) compared with 0.86 per 100 person-years (161 of 18,686) among the negatives (Table 2). The hazard ratio for ectopic pregnancy showed that women with at least 1 C. trachomatis-positive test result had a lower risk of ectopic pregnancy than women with no positive test result (crude hazard ratio, 0.40; 95% CI, 0.23–0.69); adjusted hazard ratio 0.55 (95% CI, 0.31–0.96) (Table 2).
The time to child births was evaluated in the total study cohort of 1882 women tested C. trachomatis-positive at least once and 11,811 women only having C. trachomatis-negative results (Table 1). Among the positives, 31.1% (588 of 1882) experienced such a birth compared with 29.5% (3479 of 11,811) of those who only tested negative (Table 1). The mean observation time before births was 5.0 years (data not shown). The incidence rate of births was 5.84 per 100 person-years (588 of 10,072) in the positives compared with 5.96 per 100 person-years (3479 of 58,353) in the negatives (Table 3). The time to birth was comparable in the 2 groups (crude hazard ratio, 0.98; 95% CI, 0.90–1.07; adjusted hazard ratio, 0.92; 95% CI, 0.84–1.00) (Table 3).
In this register-based study, we identified a cohort of women tested for C. trachomatis infection in various clinical settings and compared the rates of ectopic pregnancies and births of those tested positive with those only registered with negative test results. Women tested positive had a lower rate of ectopic pregnancy than those only tested negative, but their birth rates were comparable in the 2 groups.
The current study has the important advantage of an almost complete follow-up of women until their first pregnancy irrespective of their previous C. trachomatis test results. This has not previously been possible because of the timespan until potential long-term complications become manifest, and furthermore, because testing for C. trachomatis did not become generally available until the 1980s.
A number of limitations to our study must be acknowledged. Information bias with regard to having had a C. trachomatis infection cannot be avoided because approximately half of all urogenital infections are asymptomatic in the general female population,19 and these infections will have remained undetected in our study. Thus, women categorized as C. trachomatis-negative may have been misclassified because they could have had an undetected infection at another time. Information bias with regard to having ever had a C. trachomatis test is also possible because women may have been tested outside the county, but we believe that this is rather uncommon in our county-based, tax-paid healthcare system. The effect of such a bias will also represent an instance of misclassification of C. trachomatis-negatives that will tend to underestimate the consequences of having had an infection detected earlier. A C. trachomatis-positive diagnosis may also have been missed as a result of a low diagnostic efficacy of the tests used. Both Microtrak and Chlamydiazyme do not perform as well as nuclear acid amplification tests (NAATs),20 but these tests were the best available tests at that time. False-negative test results may also have occurred because of the low number of women having tests performed from both cervix and urethra, because samples obtained from the cervix only reduce the overall sensitivity of the testing by15%.21
Another problem that we unfortunately were not able to control for is confounders; data on sexual behavior and contraceptive use were not collected. Especially data on contraceptive use could have been very useful because many of the C. trachomatis-positive women can have used contraceptive pills, which may protect against upper genital tract infection. Furthermore, testing for other relevant microorganisms like Neisseria gonorrhoeae was not registered, but it should be noted that the incidence of gonorrhea was declining rapidly throughout Denmark during the study period with 10,935 recorded cases in 1984 versus 936 in 1992.22
Screening of asymptomatic individuals was not officially recommended in Denmark at the time of inclusion. However, women contacting a general practitioner for contraceptive pill prescription often have a gynecologic examination and are often tested for C. trachomatis (unpublished data from a Danish survey among general practitioners treating C. trachomatis-infected men and women; data were collected together with reference 23). Women using prescription pills may be more sexually active than the background population and hence be facing a higher risk of infection. Symptomatic women will probably also be highly represented in the cohort, and they represent a group of women who are at high risk for infection. We therefore consider our cohort to be a high-risk population. This is supported by the fact that the infection prevalence reached 10.9%, which is almost twice as high as the prevalence in the general asymptomatic population19 but corresponds well to the prevalence seen in samples obtained because of symptoms.2,24 In our study cohort, symptomatic C. trachomatis-positive women had their symptoms explained and received treatment for these symptoms, which was not the case for the C. trachomatis-negative women with symptoms. Such unexplained symptoms may have several causes, including infection with microorganisms also causing reproductive sequelae. A possible example could be Mycoplasma genitalium, which has only recently been shown to cause urethritis25 and cervicitis26 in humans, and which is also suspected to limit fertility options for women.27,28 Further research is needed to clarify the impact of other microorganisms on ectopic pregnancy and reproductive capacity.
In contrary to other studies from other countries,29 we know from an earlier Danish study performed in 1994 that nearly all cases with a positive C. trachomatis test were given relevant antibiotic treatment initiated by the general practitioner who ordered the test.23 This practice is in line with the guidelines laid down by the Danish National Board of Health in the 1980s.30 Thus, the C. trachomatis-positive group consisted of women whose infection was treated. In contrast, the inevitable occurrence of false-negative suggests that the C. trachomatis-negative group is a composite and consists both of uninfected women and women with an infection that was not treated. This may at least partly explain the lower ectopic pregnancy rate among C. trachomatis-positives compared with -negatives. This interpretation corresponds very well to results indicating that women receiving early treatment for a PID do not have an elevated risk for long-term complications.5–7 This would seem to underline the relevance of actively searching for and treating C. trachomatis infections, and it opposes the theory that immunologic factors causing tubal scarring in early stages of C. trachomatis infections should limit the benefit of early detection and treatment of the infection.31
In recent years, studies have drawn attention to the emotional consequences of C. trachomatis-positive test results; in particular women’s fear of becoming infertile as a result of the C. trachomatis infection.17 Our cohort study did not show any significant untoward effects on ectopic pregnancy and births among those infected, and these results suggest that relevant counseling of women having a C. trachomatis-positive test result should emphasize that detection and treatment of the C. trachomatis infection puts the woman in a far more advantageous situation than women who are not diagnosed and treated.
1.Gerbase AL, Rowley TR, Mertens EM. Global epidemiology of sexually transmitted diseases. Lancet 1998; 351(suppl):2–4.
2.Møller JK, Andersen B, Olesen F, Østergaard L. Reasons for Chlamydia trachomatis
testing and the associated age-specific prevalences. Scand J Lab Clin Invest 2003; 63:339–346.
3.Scholes D, Stergachis A, Heidrich FE, Andrilla H, Holmes KK, Stamm WE. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. N Engl J Med 1996; 334:1362–1366.
4.Westrom L, Joesoef R, Reynolds G, Hagdu A, Thompson SE. Pelvic inflammatory disease and fertility. A cohort study of 1,844 women with laparoscopically verified disease and 657 control women with normal laparoscopic results. Sex Transm Dis 1992; 19:185–192.
5.Haggerty CL, Ness RB, Amortegui A, et al. Endometritis does not predict reproductive morbidity after pelvic inflammatory disease. Am J Obstet Gynecol 2003; 188:141–148.
6.Hillis SD, Joesoef R, Marchbanks PA, Wasserheit JN, Cates W Jr, Westrom L. Delayed care of pelvic inflammatory disease as a risk factor for impaired fertility. Am J Obstet Gynecol 1993; 168:1503–1509.
7.Heinonen PK, Leinonen M. Fecundity and morbidity following acute pelvic inflammatory disease treated with doxycycline and metronidazole. Arch Gynecol Obstet 2003; 268:284–288.
8.Hawes LA, Gilbert GL. Seroepidemiology of Chlamydia trachomatis
infection in infertile women in Melbourne. Med J Aust 1986; 145:497–499.
9.Mehanna MT, Rizk MA, Eweiss NY, et al. Chlamydial serology among patients with tubal factor infertility and ectopic pregnancy in Alexandria, Egypt. Sex Transm Dis 1995; 22:317–321.
10.Reniers J, Collet M, Frost, Leclerc A, Ivanoff B, Meheus A. Chlamydial antibodies and tubal infertility. Int J Epidemiol 1989; 18:261–263.
11.Thejls H, Gnarpe J, Lundkvist O, Heimer G, Larsson G, Victor A. Diagnosis and prevalence of persistent chlamydia infection in infertile women: Tissue culture, direct antigen detection, and serology. Fertil Steril 1991; 55:304–310.
12.Brunham RC, Peeling R, Maclean I, Kosseim ML, Paraskevas M. Chlamydia trachomatis
-associated ectopic pregnancy: Serologic and histologic correlates. J Infect Dis 1992; 165:1076–1081.
13.Coste J, Laumon B, Bremond A, Collet P, Job SN. Sexually transmitted diseases as major causes of ectopic pregnancy: Results from a large case–control study in France. Fertil Steril 1994; 62:289–295.
14.Chaim W, Sarov B, Sarov I, Piura B, Cohen A, Insler V. Serum IgG and IgA antibodies to chlamydia in ectopic pregnancies. Contraception 1989; 40:59–71.
15.Wiesenfeld HC, Hillier SL, Krohn MA, et al. Lower genital tract infection and endometritis: Insight into subclinical pelvic inflammatory disease. Obstet Gynecol 2002; 100:456–463.
16.Johnsen S, Andersen PL, Stanek G, et al. Chlamydia antibody response in healthy volunteers immunized with nonchlamydial antigens: A randomized, double-blind, placebo-controlled study. Clin Infect Dis 2003; 36:586–591.
17.Duncan B, Hart G, Scoular A, Bigrigg A. Qualitative analysis of psychosocial impact of diagnosis of Chlamydia trachomatis
: Implications for screening. BMJ 2001; 322:195–199.
18.Andersen TF, Madsen M, Jorgensen J, Mellemkjoer L, Olsen JH. The Danish National Hospital Register. A valuable source of data for modern health sciences. Dan Med Bull 1999; 46:263–268.
19.Andersen B, Olesen F, Møller JK, Østergaard L. Population-based strategies for outreach screening of urogenital Chlamydia trachomatis
infections: A randomised controlled trial. J Infect Dis 2002; 185:252–258.
20.Østergaard L. Diagnosis of urogenital Chlamydia trachomatis
infection by use of DNA amplification. APMIS 1999; Suppl:5–36.
21.Statens Seruminstitut. EPI-NYT, uge 38: Klamydia; 1995 [in Danish], 1996.
22.Statens Seruminstitut. EPI-NYT, uge 4: Gonoré; 1991–1992 [in Danish], 1994.
23.Andersen B, Østergaard L, Nygård B, Olesen F. Urogenital Chlamydia trachomatis
infections in general practice: Diagnosis, treatment, follow-up and contact tracing. Fam Pract 1998; 15:223–228.
24.Østergaard L, Møller JK, Andersen B, Olesen F. Diagnosis of urogenital Chlamydia trachomatis
infection in women based on mailed samples obtained at home: Multipractice comparative study. BMJ 1996; 313:1186–1189.
25.Bjornelius E, Lidbrink P, Jensen JS. Mycoplasma genitalium in non-gonococcal urethritis—A study in Swedish male STD patients. Int J STD AIDS 2000; 11:292–296.
26.Manhart LE, Critchlow CW, Holmes KK, et al. Mucopurulent cervicitis and Mycoplasma genitalium
. J Infect Dis 2003; 187:650–657.
27.Clausen HF, Fedder J, Drasbek M, et al. Serological investigation of Mycoplasma genitalium
in infertile women. Hum Reprod 2001; 16:1866–1874.
28.Cohen CR, Manhart LE, Bukusi EA, et al. Association between Mycoplasma genitalium
and acute endometritis. Lancet 2002; 359:765–766.
29.Aavitsland P. Survey of the treatment of Chlamydia trachomatis
infection of the female genital tract. Acta Obstet Gynecol Scand 1992; 71:356–360.
30.Sundhedsstyrelsen. Vejledning for diagnose og behandling af seksuelt overførbare sygdomme [in Danish], 1992.
31.Shahmanesh M. Problems with non-gonococcal urethritis [Editorial]. Int J STD AIDS 1994; 5:390–399.