IN WOMEN, Chlamydia trachomatis infections can cause pelvic inflammatory disease, chronic pelvic pain, tubal infertility, and ectopic pregnancy, which constitute a major personal, social, and health service burden. These complications may be seen in 10% to 25% of infected women. 1 Since up to 70% of infections in women are asymptomatic and are therefore unlikely to be treated, 2 screening is an important public health intervention. Such programs have been successfully implemented in Sweden 3 and the United States 4,5 and are being considered in the United Kingdom. 6
The national laboratory-based surveillance system was established in 1988 and forms the backbone of infectious disease surveillance activities in Switzerland. 7 All laboratories recognized by the Swiss Federal Office of Public Health (SFOPH) must, by law, report new infections of C trachomatis on a weekly basis. C trachomatis infections are some of the most common infectious diseases reported by the laboratories, 8 and presently this is the only source of national data on chlamydial infections in Switzerland. 9
The objectives of the Sentinella Chlamydia Prevalence Study were twofold: (1) to estimate the prevalence and identify factors associated with genital chlamydial infection among women who consult their gynecologists in Switzerland and (2) to make an estimate of genital chlamydial infections among low-risk, sexually active women in Switzerland and to compare this estimate with national laboratory reports of C trachomatis. The first objective was to assess the feasibility of screening women consulting private gynecologists for genital chlamydia infections in Switzerland, and the second was to estimate the sensitivity of laboratory reports of C trachomatis (sensitivity at the level of case-reporting refers to the proportion of cases of a disease or health event detected by a surveillance system 10).
The study was performed between December 27, 1997, and September 4, 1998. It was carried out within the framework of the Swiss Sentinel Surveillance Network of Gynecologists (SSSNG), which is a network of gynecologists working in private practices. Participation in the surveillance system is voluntary and unpaid, and the network handles an estimated 2.5% to 3.5% of private practice gynecological consultations in Switzerland. 11
Gynecologists in the SSSNG sampled two groups of women aged less than 35 years: those having first consultations for a pregnancy and sexually active women having a routine check-up. A cervical swab specimen for C trachomatis was first obtained from a pregnant woman and then from the next woman consulting the practice for a routine check-up. The gynecologist then waited for the next pregnancy consultation before sampling another check-up woman. The cervical swab specimens were obtained after the women granted informed consent and the gynecologists completed an anonymous questionnaire for each woman, covering simple sociodemographic information (e.g., age, nationality) and clinical features (e.g., urogenital symptoms). The questionnaires were sent to the Sentinella Chlamydia Prevalence Study surveillance unit at the SFOPH at the end of each week.
The cervical swabs were sent (by post) to the Department of Medical Microbiology at the University of Zurich, where they were tested with the plasmid-based ligase chain reaction (LCR) assay (LCX; Abbott Laboratories, Chicago, IL), according to the manufacturer's instructions. The Department of Medical Microbiology sent the laboratory result to the gynecologist by post or fax, depending on the how he or she preferred to receive the result. A copy of the laboratory report was rendered anonymous by removing the patient's name and sent to the Sentinella Chlamydia Prevalence Study surveillance unit. The laboratory test results were linked to the questionnaires by the study questionnaire numbers.
We assumed that the prevalence of C trachomatis among the check-up women would be a rough estimate of the prevalence of C trachomatis among low-risk, sexually active women in Switzerland. This was based on the fact that women in Switzerland have open access to private gynecologists and they use this service. In the Swiss Health Survey carried out in 1997, 12 50% of women (all ages) reported that they had consulted a gynecologist in the previous 12 months (compared with 65% who had consulted a general practitioner). In the age group of 15 to 24 years, the percentage was 50%, and in the age group of 24 to 34 years it was 68%. 12
The prevalences observed among low-risk, sexually active women were applied to the Swiss population statistics 13 to provide a conservative estimate of the total number of female genital chlamydial infections in Switzerland. The estimates are conservative, as they exclude high-risk women (e.g., women attending public family planning clinics 14,15 or sexually transmitted disease [STD] clinics 15). Independent predictors of C trachomatis among the check-up women were sought to refine the estimates. For example, if significantly higher prevalences were observed among non-Swiss women, the extrapolations would be carried out separately for Swiss and non-Swiss women.
The univariate analyses were performed and the chi-square and chi-square for trend values were calculated with use of EPI-INFO (version 5.0; Centers for Disease Control and Prevention, Atlanta, GA). The multivariate analyses were performed with EGRET version 0.25.6. Independent risk factors of chlamydial infections were assessed by means of stepwise logistic regression.
Prevalence of C trachomatis
Cervical swab specimens were collected from 1751 women and tested by LCR. Of these, 119 were excluded from the analysis because they did not meet the inclusion criteria (mainly women who were older than 34 years) and 43 were excluded because they were undergoing antibiotic treatment (in the previous 2 weeks). The analysis was therefore based on a total of 1589 cervical swab specimens (91%): 817 from pregnant women and 772 from check-up women. The cervical swabs were collected by 36 gynecologists working in 15 of the 26 cantons in Switzerland. Each gynecologist collected an average of 44 cervical swabs (range, 1–190).
Pregnant women were more likely to be older (median age of 29 years, versus 27 among check-up women), to be in a stable relationship (98.9% versus 89.0%;P = 0.000; data not shown), to be foreign (21.2% versus 12.6%;P = 0.000), and not to have urogenital symptoms (5.3% versus 8.3%;P = 0.02). The prevalence of genital chlamydial infection (Table 1) was twice as high among check-up women (2.9%; 95% CI: 1.8–4.2%) than among pregnant women (1.4%; 95% CI: 0.7–2.3%).
The prevalences in different subgroups did not vary significantly for any of the variables analyzed, except the presence of urogenital symptoms (Table 1). Women consulting their gynecologist with urogenital symptoms were significantly more likely to be infected with C trachomatis than those without these symptoms, in both groups of women. Genitourinary symptoms were reported in 2 of the 11 cases of genital chlamydial infection among pregnant women (18%) and in 6 of the 22 cases among check-up women (27%).
Urogenital symptoms were not defined in the prevalence study, and the definition was left to the discretion of the gynecologists in the SSSNG. The gynecologists were asked to list these symptoms and, where this information is available, the most common symptoms stated were vaginal discharge, in 23% (10) of 43 women with urogenital symptoms who were pregnant and 22% (14) of 64 check-up women, and candidiasis (28% [12/43] among pregnant women and 17% [11/64] among check-up women).
The second objective of the Swiss Chlamydia Prevalence Study was to make a conservative estimate of the total number of genital chlamydial infections among low-risk, sexually active women in Switzerland. The multivariate analysis among check-up women did not reveal any population group for which the prevalences should be stratified, for example, Swiss and non-Swiss (Table 1). Since no infections were found in those aged <20 years (data not shown) and we expected higher prevalences in this age group on the basis of previous studies 14,15 (see discussion), we excluded this group from the extrapolations. Prevalences varied by age among check-up women who were older than 19 years (P = 0.07; see the last footnote in Table 1); therefore, we performed the extrapolations by 5-year age groups with use of the Swiss census statistics 13 (Table 2). Our conservative estimate of the total number of genital chlamydial infections among women aged 20 to 34 years was 24,400 (95% CI: 14,300–34,300) in 1998.
Performing the analysis with no age stratification (in other words, applying the prevalence seen among check-up women aged 20–34 years to all women in Switzerland in this age group) hardly modified the estimate (24,300; 95% CI: 15,200–36,400). National laboratory reports to the SFOPH of C trachomatis among women aged 20 to 34 years in 1998 accounted for only 4.7% of the estimated number of genital infections in Switzerland (Table 2).
A number of C trachomatis prevalence studies have been carried out among women in Switzerland. These studies have been characterized by the highly selective nature of the study populations. Two studies were carried out among women attending family planning clinics, and observed prevalences were 18.5% in Lausanne (1987) 14 and 9.6% in Zurich (1995–1997). 15 Another study, of men and women consulting a policlinic of dermatology, was carried out in Zurich (1995–1996) and showed a prevalence of 13.7% among women. 15
An attempt to estimate the prevalence of C trachomatis in the general population in the Zurich agglomeration was made in July 1995. 15 A total of 2000 residents were randomly sampled and sent a letter describing the objectives of the study, a letter of consent, a detailed questionnaire, and the request to come to the University Hospital to give a urine sample. A total of 52 persons voluntarily gave a urine sample, and another 48 did so following a telephone recall of 500 nonparticipants. A total of 100 urine samples were therefore collected (5% of those contacted), and no one tested positive for C trachomatis. 15 This study highlights the difficulty of trying to estimate the prevalence of C trachomatis in the general population and the need to explore indirect approaches, such as the Sentinella Chlamydia Prevalence Study, to obtain this information.
Very few prevalence studies have been carried out in Western Europe among women consulting their gynecologists, and the results from other studies are difficult to compare because of the different methodologies and clinical settings. A study carried out in Berlin among women aged 20 to 39 years consulting their gynecologists for cervical smear testing or to obtain a prescription for oral contraceptives showed a prevalence of 3.6% (n = 5022). 16 Another, carried out by 46 gynecologists in the greater Paris area, who screened all consecutive female attendees during 1 week, showed a prevalence of 0.8% (n = 1893). 17 A number of studies have been carried out in the United Kingdom among women consulting their general practitioners. One in London among women aged 18 to 35 years consulting their general practitioners for cervical smear testing or a “young well woman” check-up showed a prevalence of 2.6% (n = 890). 1 Another, in Scotland, showed a prevalence of 3.5% among patients consulting 619 general practitioners in the Lothian region for genital swab testing for chlamydia. 18
Previous study reports have suggested that screening for C trachomatis becomes cost-effective at a prevalence of 6% or more. 19–22 The Sentinella Chlamydia Prevalence Study showed such prevalences only among check-up women with urogenital symptoms (9.4%;Table 1). From a purely cost-effective perspective, our study results support systematic C trachomatis screening only for check-up women aged less than 35 years who have urogenital symptoms.
In an attempt to identify a screening strategy that captures more of the C trachomatis infections (only 6 [27%] of the 22 infections among check-up women would have been identified if screening were limited to urogenital symptoms), we explored other selective screening criteria. The only criteria that identified a high proportion of infections (>80%) were those that led to testing of more than 60% of the women in each group. Among the pregnant women, this criterion was all women aged less than 30 years (82% of infections and 61% of women), and among check-up women, it was all women who had urogenital symptoms and were aged less than 30 years (81% of infections and 69% of women).
Considering (1) the objectives of the Sentinella Chlamydia Prevalence Study were not to identify selective screening criteria, (2) our study probably omitted important screening questions (see below), and (3) screening recommendations should be based on rigorous cost-effectiveness studies, we feel that this issue requires further attention. A new study looking specifically at the question of selective screening in this population should explore variables that we did not include in our study. The chlamydia prevalence studies carried out in England 1 and France 17 highlighted the importance of questions concerning “sexual partners in the previous 12 months” (whether the women had 2 or more sexual partners in the previous 12 months 1 and whether the women had a new sexual partner in the previous 12 months 17). Another important question that could be explored is whether the women have had previous treatment for a sexually transmitted disease.
Our estimate of the sensitivity of laboratory reports of C trachomatis among women aged 20 to 34 years (4.7%) is based on a fixed numerator (the total number of laboratory reports) and an estimated denominator (the extrapolation estimate). A number of factors can affect these two numbers.
The numerator is based on laboratory reports of C trachomatis to the SFOPH by recognized laboratories in Switzerland. A survey of all laboratories in Switzerland—recognized and nonrecognized, public and private—showed that laboratory reports of C trachomatis to the SFOPH represented roughly 70% of all positive laboratory tests for C trachomatis in 1994. 23 We have no reason to believe that this proportion changed 4 years later, and the sensitivity estimate would probably not increase greatly if all laboratories in Switzerland were to report C trachomatis to the SFOPH (from 4.7% to 6.7%). The laboratory reports of C trachomatis include genital, urethral, eye, and lung infections; the total number of genital infections is therefore lower than the 1150 reports of infections in females aged 20 to 34 years received by the SFOPH in 1998. Correcting the sensitivity estimate for this factor would produce a lower percentage.
The estimate of the denominator (the total number of genital chlamydial infections among women aged 20–34 years) is based on check-up women who consult a private gynecologist. A number of selection biases may have affected our prevalence study results. The Sentinella Chlamydia Prevalence Study excluded women who consulted public hospitals (e.g., women's hospitals) and private clinics/hospitals. There was also an important underrepresentation of foreign women (only 12.6% of the check-up women were foreign, compared with 26.0% in the general population aged 20–34 13), a group that had a higher, though not significantly higher, prevalence. Also, we had data from all of the major urban cantons but missed women from many of the rural cantons in Switzerland. Overall, we feel that the prevalences in the check-up group probably underestimated the true prevalence of C trachomatis in women consulting a gynecologist in Switzerland, mainly because of the lack of data from public hospitals and the deficit of foreign women.
The overall impact of these different factors on the sensitivity estimate for genital chlamydial infections is that our estimate is probably too low. The numerator is too high because it includes all types of chlamydial infections, and the denominator is too low because it is based on a low-risk population. Overall, the sensitivity of laboratory reports of C trachomatis infections for genital chlamydial infections among women aged 20 to 34 years is therefore probably less than 4.7% in Switzerland. These findings are important for the evaluation of STD surveillance systems in Switzerland, as they demonstrate the limitations of laboratory reports in assessing the frequency of genital chlamydial infections among women and highlight the importance of using prevalence studies to measure the frequency of a predominantly asymptomatic infection.
1. Grun L, Tassano-Smith J, Carder C, et al. Comparison of two methods of screening for genital chlamydial infection in women attending in general practice: cross sectional survey. BMJ 1997; 315: 226–230.
2. Schachter J, Stoner M, Moncada J. Screening for chlamydial infections in women attending family planning clinics. West J Med 1983; 138: 375–379.
3. Egger M, Low N, Davey Smith G, Lindblom B, Herrmann B. Screening for chlamydial infections and the risk of ectopic pregnancy in a county in Sweden: ecological analysis. BMJ 1998; 316: 1776–1780.
4. Mertz KJ, Levine WC, Mosure DJ, Berman SM, Dorian K. Trends in the prevalence of chlamydial infections: the impact of community-wide testing. Sex Transm Dis 1997; 24: 169–175.
5. Stamm WE. Expanding efforts to prevent chlamydial infection. N Engl J Med 1998; 339: 768–770.
6. CMO's Expert Advisory Group. Chlamydia trachomatis. London: Department of Health, 1998.
7. Stürchler D. Meldung von Infektionskrankheiten in der Schweiz. Schweizerische Ärztezeitung 1999; 80: 715–717.
8. Office federal de la sante publique. Maladies infectieuses en Suisse 1998. Bern: Swiss Federal Office of Public Health, 2000.
9. Paget WJ, Zimmermann H-P. Surveillance of sexually transmitted diseases in Switzerland, 1974–1994: evidence of declining trends in gonorrhoea and syphilis. Soz Präventivmed 1997; 42: 30–36.
10. Centers for Disease Control and Prevention. Updated guidelines for evaluating public health surveillance systems: recommendations from the guidelines working group. MMWR Morb Mortal Wkly Rep 2001; 50 (RR-13):1–35.
11. Matter HC, Birrer A, Caesar J, Paget WJ, Barthe M, Schumacher R. Le système de déclaration Sentinella en Suisse: aspects méthodologiques concernant la période 1998. In: Sentinella 1998: Annual Report of the Swiss Sentinel Surveillance Network. Bern: Swiss Federal Office of Public Health, 2000.
12. Calmonte R, Spuhler T, Weiss W. Gesundheit und Gesundheitsverhalten in der Schweiz 1997. Neuchâtel: Office fédéral de la statistique, 2000.
13. Office fédéral de la statistique. Annuaire statistique de la Suisse 1998. Zürich: Neue Zürcher Zeitung, 1998.
14. Nguyen Duy C, Bonanomi Schumacher S, Broel-Schneider C, et al. Infection cervicale àChlamydia trachomatis
dans un centre de planning famillial: prévalence, analyse des facteurs de risque, modèle de prédiction. J Gynecol Obstet Biol Reprod 1989; 18: 977–987.
15. Lee C-Y. Die urogenitale Chlamydieninfektion in der Schweiz—eine Pilotstudie. Doctoral dissertation. Zürich: University of Zürich Medical School, 1997.
16. Koch J, Kirshner W, Schäfer A. Bestimmung der Prävalenz genitaler HPV- und Chlamydia-trachomatis
-Infektionen in einem repräsentativen Querschnitt der weiblichen Normalbevölkerung in Berlin. Infektionsepidemiologische Forschung 1997; II: 1–7.
17. Warszawski J, Meyer L, Weber P. Criteria for selective screening of cervical Chlamydia trachomatis
infections in women attending private gynaecology practices. Eur J Obstet Gynaecol Reprod Biol 1999; 86: 5–10.
18. Ross JDC, Sutherland S, Coia J. Genital Chlamydia trachomatis
infections in primary care. BMJ 1996; 313: 1192–1193.
19. Genç; M, Mårdh P-A. A cost-effectiveness analysis of screening and treatment for Chlamydia trachomatis
infection in asymptomatic women. Ann Intern Med 1996; 124: 1–7.
20. Trachtenberg AL, Wahington E, Halldonson S. A cost based decision analysis for chlamydia screening in California family planning clinics. Obstet Gynecol 1988; 71: 101–108.
21. Estany A, Todd M, Vasquez M, McLaren R. Early detection of genital chlamydial infection in women: an economic evaluation. Sex Transm Dis 1989; 16: 21–27.
22. Phillips RS. Should tests for Chlamydia trachomatis
cervical infection be done during routine gynaecologic visits? Cost and analysis. Ann Intern Med 1987; 107: 188–194.
23. Paget WJ. Laboratory Evaluation Study of 1994:Chlamydia trachomatis
, Neisseria gonorrhoeae
, Treponema pallidum
, Mycobacterium tuberculosis
. Internal Report. Bern: Swiss Federal Office of Public Health, 1997.