Opportunistic testing for urogenital infection with Chlamydia trachomatis in south-western Switzerland , 2012 : a feasibility study

F Bally (frank.bally@hopitalvs.ch)1,2,3, A Quach3,4, G Greub5, K Jaton5, C Petignat6, C Ambord2, J Fellay7, E Masserey6, B Spencer8 1. Institut Central (Hôpital du Valais), Service des maladies infectieuses, Sion, Switzerland 2. Service de la santé publique du canton du Valais, Sion, Switzerland 3. These authors contributed equally to the manuscript 4. Profa Consultation de santé sexuelle planning familial, Renens, Switzerland 5. Institute of Microbiology, University Hospital of Lausanne, Lausanne, Switzerland 6. Service de la santé publique du canton de Vaud, Lausanne, Switzerland 7. Centre Sexualité-Information-Prévention-Education (SIPE), Sion, Switzerland 8. University Institute of Social and Preventive Medicine (IUMSP), Lausanne, Switzerland


Introduction
Chlamydia trachomatis is a frequent cause of sexually transmitted urogenital infections [1].Carriers with asymptomatic infection are a difficult to reach reservoir promoting transmission to their sexual partners [2].Complications, although rarely life threatening, can be substantial, especially for women.They include pelvic inflammatory disease, chronic abdominal pain, ectopic pregnancy, tubal sterility [2] and possibly a higher risk for adverse pregnancy outcomes [3,4].
Rates of Chlamydia-related complications in a given population correlate with the prevalence of chlamydial infection [5].Treatment of urogenital infections caused by C. trachomatis can prevent complications, at least in the short term [6,7].The pooled risk ratio for all-cause pelvic inflammatory disease after one year of followup in women invited to have C. trachomatis screening in four randomised controlled trials was 0.64 (95% confidence interval (CI): 0.45-0.90)[2].Complications may occur despite regular screening at fixed intervals because of infection after treatment or during screening intervals [8,9].It has also been hypothesised that early treatment may impede development of immunity and favour future re-infection [10,11].Following a decline in the late 1980s and early 1990s, laboratory notifications of infections with C. trachomatis in Switzerland have more than quadrupled since 2000 [1,12].In 2003, most infections were diagnosed by gynaecologists, hospital services and primary healthcare physicians [13].One study, published in 1989, found a positive culture rate of 18% in 600 women aged 18 to 55 years at a sexual health centre in Lausanne [14].Half of these women (49%) were symptomatic.More recent studies in Switzerland using PCR testing found lower rates: In 1998, 1% of 817 pregnant women and 2.8% of 772 other sexually active women were found to be PCR-positive for Chlamydia by their gynaecologist [15].In 2006 and 2007, 1.2% of 517 male Swiss military recruits with a mean age of 20 years were found by PCR to be infected [16], as were 7.3% of 386 healthy pregnant women in the period from 2006 to 2009 [4].
The European Centre for Disease Prevention and Control (ECDC) recommends implementation of C. trachomatis control using a strategy of four levels: primary prevention, case management, opportunistic testing and systematic screening [2,17].In Switzerland, a national programme for primary prevention of human immunodeficiency virus (HIV) infection was started in 1987, and subsequently widened in 2011 to all sexually transmitted infections (STI) [18].A national guideline for case management of STI including C. trachomatis was published in 2011 [19], but no recommendations exist for testing.A C. trachomatis test with administration fees costs CHF 119 (EUR 111), not including any medical consultation fees.These costs are reimbursed by basic insurance when the yearly medical costs exceed CHF 300 (EUR 281).The young and healthy without other health expenses therefore pay screening costs directly.
This study explores the feasibility of opportunistic testing for C. trachomatis control, the third level in the ECDC recommendations.From a public health perspective, feasibility should be examined at all stages of programme implementation, from societal to individual level.These may be conceptualised as political acceptance, provider compliance, target population acceptance, and user compliance.We report on feasibility at all of these levels.

Methods
The study was conducted in two cantons with a combined population of 1 million, situated in the southwestern part of Switzerland: Valais and Vaud.The capital cities of Vaud and Valais, Lausanne and Sion, have 142,000 and 42,000 inhabitants, respectively.Both cantons have rural districts, some of them extending into partially remote alpine valleys.Most districts are French speaking; German is spoken in the eastern districts of Valais (about one in twelve of the total study population).
Free C. trachomatis testing in first-void urine or, for women according to personal preference, self-applied vaginal swabs, was offered from February to August 2012 to all persons younger than 30 years in public health services representative of the whole territory.These included all centres of two public cantonal sexual health networks (eight in Vaud, including seven with an on-site physician; five in Valais, none with on-site physicians) and, for comparison, two infectious disease (ID) outpatient clinics (Sion and Visp, Valais).As the number of available tests was restricted by the allocated study budget, the recruitment period was shortened in centres with high testing activity in order to allow testing in centres with lower throughput.Every candidate, defined as a female or male person visiting a screening centre or clinic, was given an invitation (invited candidates) for screening together with an information sheet about C. trachomatis.Individuals who had never had sexual intercourse or who had been screened for urogenital C. trachomatis infection less than three months previously were excluded.Participation was confirmed by written consent.Consenting participants (Figure 1) were given a questionnaire on demographics and sexual behaviour and a self-sampling kit with illustrated instructions on how to sample urine and, for women, how to take a vaginal swab.
The urine and vaginal swab samples of tested participants were collected at the screening centre and centrally analysed by PCR.In Valais, a commercial kit (Roche Diagnostics, Switzerland) was used by the Central Institute laboratory for molecular biology, Sion (Institut Central Hôpital du Valais; ICHV).In Vaud, a validated in-house C. trachomatis PCR [20] was used by the Institute of Microbiology of the University of Lausanne (IMUL).
According to the participant's choice, treatment for Chlamydia was organised with their primary care physician or gynaecologist, at the screening centre or at the nearest centre with an on-site physician.Partner notification was recorded for each infection.Screening for other STI and preventive counselling were left to the discretion of the treating physician.A control visit, involving a second free test for C. trachomatis, was scheduled for cases with documented infection six months after treatment.Each screening centre had to document for each candidate all the steps up to either a negative screening result or, in the case of a positive result, the negative control test result six months after treatment.The first step not fulfilled was noted as the point of drop-out.
The study protocol was submitted and accepted by the ethical committee in each participating canton (Valais: no.CCVEM 023/11; Vaud: no.281/11).The study was supervised by a committee of representatives of the participating screening centres, the public health authorities and the testing laboratories.It also included a research specialist from the University Institute of Social and Preventive Medicine, Lausanne.
Descriptive statistics and comparisons (Fisher's exact test for 2x2 tables (proportions), chi-square test for other tables and Kruskal-Wallis test for continuous variables) and other calculations were produced with open source R, version 3.0.3[21].Confidence intervals for proportions were calculated with the asymptotic definition for confidence limits on a single proportion using the Central Limit Theorem (binom.testfunction in package binom).

Results
Results regarding each successive level of the study flow are summarised in Figure 1.
Provider compliance, assessed by rates of screening invitations, and target population acceptance, assessed by rates of participation, could be monitored in 14 of the 15 screening centres, totalling 2,995 candidates between February and December 2012.One centre did not consistently distinguish between noninvitation and non-participation and was therefore excluded from this part of the analysis.This centre was  a One of 15 centres did not consistently distinguish between non-invitation and non-participation and was excluded from the inclusion part of analysis.This centre provided an additional 314 consenting participants included in screening and follow up.b Control test by PCR at 6 months after treatment.c Negative control test by PCR at six months after treatment.
A mean of 149 consenting participants (125 women, 24 men) were tested per week when all centres were open.Extrapolated over 52 weeks, assuming access to testing under the same conditions and 82% of the population sexually active (personal communication: Lociciro S, Spencer B, July 2014), 3.7% (6.7% of women, 1.2% of men) of the sexually active population aged between 12 and 29 years in 2012 in the study region [22] would have been tested.
Infection rates were similar for different educational levels, districts of residence, sizes of population of the town of residence (≥ 10,000 vs < 10,000 inhabitants) and country of birth (Switzerland vs other).No infection was found in 66 female participants indicating no sexual partner in the last six months.The four districts with only one, seven, 13 and 39 tested participants reported zero infections.Infection rates were higher for sexual health patients (including general counselling, gynaecology checks, pregnancy, STI screening) than for patients in travel clinics and other sources, with infection rates of 127 of 2,122 (6.0%, 95% CI 5.0%-7.0%),two of 177 (1.1%, 95% CI 0-2.7%) and five of 156 (3.2%, 95% CI 0.4-6.0%),respectively.Patients in travel clinics were older than sexual health patients at screening (median age: 24.8 vs 20.6 years, p < 0.0001) and at their first sexual intercourse (median age: 17 vs 16 years, p < 0.01).Of all 134 infections, 63 (47%) were found in participants indicating one sexual partner in the last six months.For both sexes, the infection rate increased with an increasing number of sexual partners in the six months before screening (Figure 4), with 19 infections in 169 participants (11%, 95% CI: 6.5-16.0)indicating more than three sexual partners in the last six months.Of 140 (5.7%)participants reporting symptoms of STI, 16 (11%) had a positive screening test.
In 2012, 974 Chlamydia infections were notified by laboratories in the study region (not including those found in this study).Extrapolating the study experience over one year, 375 infections in women (infection rate 5.9%) and 47 in men (infection rate 3.9%) could have been diagnosed by the study test centres, increasing the study region's total number of diagnosed infections by 422 infections (43%).

Discussion
Independently of the on-site presence of a physician, testing for urogenital infection with C. trachomatis in sexually active adults younger than 30 years using selfapplied urine samples, or, for women, vaginal swabs, proved technically feasible in the two Swiss cantons under study.When offered at no cost, C. trachomatis testing proved acceptable overall, despite the fact that almost half of participants had never previously heard of C. trachomatis infection (Table 2).Only 31% of all participants were consulting for HIV/STI-screening, mostly anonymous HIV-screening.C. trachomatis screening can be proposed successfully in situations other than STI screening, particularly those related to sexual health, but also in those a priori unrelated to sexual health such as travel counselling, with acceptance rates that were not much lower (78%), independently of factors such as age and sex.The German-speaking region, a more secluded and rural mountain community, had not only a lower acceptance rate, but significantly less consultation activity.This may be best explained by cultural factors resulting in less demand and geographical factors impeding easy access to testing facilities.With the exception of one walk-in facility, it was possible to integrate Chlamydia screening into the centres' daily workload without adjustment in the workforce or increased consultation times.The information sheet and illustrated instructions for sampling proved helpful.C. trachomatis PCRs exhibit similar positivity rates in urine and vaginal swabs, but urine sampling was preferred over vaginal swabs by ca 80% of women.
In this study, infection rates varied between 1% and 11%, depending on already known risk factors, and were not substantially different from infection rates reported in Switzerland and Europe [4,15,16].
Population-based studies in European Union Member States report infection rates between < 1% and 10% for women and between < 1% and 6% for men, depending on country and characteristics of the study population [2].The National Chlamydia Screening Programme (NCSP) in the United Kingdom (UK) reported an infection rate of 7.7% in 2012 [23] and a Dutch pilot study (2008-2011) a rate of 4.3% [23,24].Infection rates for men at similar risk are consistently lower than for women of the same age and level of risk [2,23,24].Our study did not find different infection rates in individuals born in Switzerland vs those born elsewhere.
Nearly half of the female participants were younger than 20 years and the study participants had a profile of low to medium risk, with a median of one sex partner in the six months before screening.Half of all infections were therefore diagnosed in low-risk participants with only one sexual partner in the past six months.Although the questionnaire identified factors representing a relatively higher risk, such as presence of symptoms (11% positivity rate), being female and in the peak age group (10%) or having more than three sex partners in the six months before screening (11%), the only characteristic specific to individuals with no risk of infection was having had no sex partner in the past six months (having been sexually active before).Risk-based selection algorithms aiming at improved cost effectiveness [25] may therefore miss a substantial number of infections occurring in parts of the population associated with lower risk.

Limitations and challenges
Despite C. trachomatis screening proving to be technically feasible and reliable and treatment being simple and affordable, important obstacles remain.Despite near universal insurance coverage, access to screening for C. trachomatis in Switzerland is limited owing to a mandatory minimal yearly participation of CHF 300 (EUR 280).The high cost of a C. trachomatis test needs to be reviewed in order to allow affordable access, especially for adolescents.Men proved to be more difficult to reach in this context, constituting only 23% of the sample.Male participants were older than female participants and the proportion younger than 20 years was significantly lower.In the NCSP, men were more likely to order self-applied tests on the Internet than to visit a clinic for testing, and the number of tests ordered in this way increased from <1% to 6% of all tests between 2006 and 2010 [26].A screening programme in 13 schools in New Orleans, United States, in 1995 to 2005 showed up to 49% repeated testing in male students (with parental consent) [27].Of men aged 18 to 35 years in a survey in the UK, 75% had seen their family doctor in the last year, without relevant differences between different age groups among the 18 to 35-year-olds, providing general practitioners with occasions for opportunistic STI screening [28].In the NCSP, 9% of male 15 to 24 years-olds were tested by their general practitioner.Extrapolated over one year, the centres in our study alone would have tested a small proportion, ca 4%, of the 12 to 29 years-olds in the study region [1].Such testing activity would thus have little impact on Chlamydia transmission or its prevalence on a population level.In Switzerland, gynaecologists, hospitals services and primary care physicians notified most of all notified C. trachomatis infections in 2003 [13], but those healthcare providers were not included in this study.School-initiated home testing in post-obligatory schools (age 16 years and upwards) was initially intended.The school authorities in one canton declined participation, which shows a limitation on political grounds.
Nineteen per cent of all partners of tested participants with a positive result could be notified.Most partner notifications were only documented by asking the study participants.Whether sex partners actually received treatment was not assessed.Ascertaining if all partners are treated is a difficult challenge [29] and would also have been difficult in this study.

Figure 4
Figure 4Chlamydia trachomatis infection rate by number of sex partners in the six months before screening, Switzerland, 2012 (n = 2,438; 17 missing answers)

Table 1
Reasons given for declining Chlamydia trachomatis screening by 269 of 384 persons invited in 14 of 15 screening centres, Switzerland, 2012 (n = 269)

Table 2
Characteristics of tested participants and questionnaire answers, Chlamydia trachomatis screening, Switzerland, 2012 (n = 2,455) Valais: travellers coming for vaccination (infectious disease clinic in Sion).b Valais: mostly patients treated for other infectious diseases such as HIV and hepatitis C (infectious disease clinic in Sion).
NA: not applicable; STI: sexually transmitted infection.Statistic tests (comparisons between the two cantons): for proportions (2x2): Fisher's exact test; for continuous data: Kruskal-Wallis.p values are not corrected for multiple testing.Totals can be lower than those in the column header because of missing answers or counts in subpopulations.a this trend corresponds to an increasing incidence or other factors such as increased screening or an increased notification rate.Based on our results four main statements can be made to inform the public health authorities of Valais and Vaud regarding preventive measures for urogenital C. trachomatis infection and its complications.Firstly, C. trachomatis is present in the study region and therefore screening and efficient treatment would be desirable to prevent complications, no less than in other countries with similar infection rates.Secondly, as this study shows that C. trachomatis screening in existing sexual health centres in south-western Switzerland is technically feasible, these screening services can also be used for epidemiological investigation.Thirdly, C. trachomatis testing at low and affordable cost could promote use by those at risk.Finally, more screening opportunities need to be created, especially for difficult to reach populations such as men or people living in regions with difficult access for geographical reasons, and drop-outs during the screening and follow-up process need to be decreased.Georges Dupuis, former Head of public health in Valais, initiated the planning of this study.We are very grateful to the staff all screening centres for their enthusiastic participation in this study.We thank Mark Chung from the Melbourne Sexual Health Centre for providing us with illustrated instructions on how to take samples which could be adapted to local needs.Thanks go to Helen Di Lallo for proof-reading the manuscript.Funding: The study received two grants from public health services, one for each canton.The Institut de Microbiologie Universitaire in Lausanne executed the Chlamydia tests at a lower than usual rate for Vaud and Roche Diagnostics, Switzerland, provided commercial Chlamydia PCR kits at a preferred price for Valais.Study organisation: Mandate Valais: the Service des maladies infectieuses, Institut Central (Hôpital du Valais -ICHV), was mandated by the Service de la santé publique de l'Etat du Valais.Vaud: the Fondation Profa, Consultations de santé sexuelle, was mandated by the Service de la santé publique de l'Etat de Vaud.Study committee (in alphabetical order): Valais: Christian Ambord, MD (Médecin cantonal, Service de la santé publique du Valais, Sion).Frank Bally, MD (Médecin-chef, Institut Central de l'Hôpital du Valais (ICHV) et Service de la santé publique du Valais, Sion), principal investigator for Valais and study administrator.Jacqueline Fellay (conseillère en santé sexuelle et reproductive, centre SIPE de Sion, Sion).Vaud: Gilbert Greub, Prof., MD PhD (Institute of Microbiology, University Hospital of Lausanne, Lausanne).Katia Jaton, PhD (Institute of Microbiology, University Hospital of Lausanne, Lausanne).Eric Masserey, MD (Médecin cantonal adjoint pour les maladies transmissibles, Service de la santé publique, Lausanne).Christiane Petignat, MD (Médecin associée, responsable Unité HPCI, CHUV et Service de la santé C. trachomatis notifications inSwitzerland have increased from 2,123 in 1999 to 9,701 in 2014.It remains unknown if