Evaluation of screening for Chlamydia trachomatis among young women in primary health care services in Manaus , Amazonas State , Brazil

Screening for Chlamydia trachomatis is not routinely offered to young asymptomatic women in Brazil. This study evaluated the performance, usefulness, and operational suitability of the Digene Hybrid Capture II (HCII) CT-ID DNA-test as an opportunistic screening tool to detect C. trachomatis in the public health system in Manaus, Amazonas State. Women aged 14-25 years who attended primary health care services were interviewed and one cervical specimen was collected during cytological screening. The HCII CT test was evaluated for its ability to detect the presence of C. trachomatis and against real-time PCR (q-PCR) in a subset of samples. Operational performance was assessed through interviews with providers and patients. Overall, 1,187 women were screened, and 1,169 had a HCII CT-ID test result (292 of these were also tested by q-PCR). Of those, 13.1% (n = 153) were positive. The sensitivity, specificity, positive and negative predictive values of HCII CT were 72.3% (95%CI: 65.4-78.6), 91.3% (95%CI: 84.195.9), 93.8% (95%CI: 88.5-97.1), and 64.4% (95%CI: 56.0-72.1), respectively. Sample collection caused discomfort in 19.7% of women. Among health professionals (n = 52), the main barriers reported included positive cases who did not return for results (56.4%), unwillingness to screen without an appointment (45.1%), and increase in their workload (38.8%). HCII CT-ID identified a high proportion of C. trachomatis cases among young women in Manaus. However, its moderate sensitivity limits its use as an opportunistic screening tool in primary health care settings in Manaus. Screening was well accepted although the barriers we identified, especially among health professionals, challenge screening detection and treatment efforts. Chlamydia trachomatis; Mass Screening; Women’s Health Correspondence


Introduction
Chlamydia trachomatis is the most commonly diagnosed sexually transmitted bacterial infection (STI) worldwide 1,2 .Several countries such as The Netherlands and the United Kingdom have witnessed an increase in C. trachomatis case reports over the last decade 3,4 , although the overall burden has decreased in other countries including the United States 5 .
Genital infection with C. trachomatis is asymptomatic in 50-88% of Colombian women 6 and is most common in young women 7 .Untreated infections can cause pelvic inflammatory disease (PID) 8 , ectopic pregnancy 9 , and subfertility 10 .Risk factors for C. trachomatis infection include young age, having more than one sexual partner, and recent change of partner 11,12 .
Given the frequent asymptomatic nature of the infection and the importance of early treatment to reduce transmission and prevent C. trachomatis-related morbidity, many developed countries are offering opportunistic screening to all sexually active people under the age of 25 years 1,13 , except the US where screening focuses only on women younger than 25 years 14 .Evidence from clinical trials suggests that screening is effective in reducing the incidence of PID while a large nonrandomized cohort found no benefit of offering screening in reducing sequelae in women 15 .
However, screening implementation might be hampered by operational and technical difficulties 16 .Barriers to health practitioners offering a test include lack of time and reluctance to raise sexual health issues within general practice 17 .Young people are willing to accept screening although low risk perception, poor health seeking behavior mainly due to the asymptomatic nature of the infection, poor understanding of what testing involves, and embarrassment remain important obstacles for screening uptake 18,19 .
In Brazil, although data on C. trachomatis population prevalence is scarce, a study conducted in six cities estimated that 9.4% of women in the general population were infected 20 and nationwide, parturient women under 25 years of age showed a prevalence of infection at 9.8% 11 .C. trachomatis screening is not systematically offered at health services.The Digene Hybrid Capture II DNA test (Qiagen, Mississauga, Canada) to identify C. trachomatis (HCII CT-ID) is the only molecular test approved system for C. trachomatis screening in the Brazilian public health system.This nucleic acid hybridization assay is no longer recommended by the US Centers for Disease Control and Prevention (CDC) for routine use based on performance 21 .This test has shown sensitivity ranging from 93.8%-97.7%and specificity ranging from 95.9%-100.0%compared to those of culture 22,23,24,25 .Cell culture was considered the gold standard for detection of Chlamydia due to specificity as high as 100% 21 .However, culture presents disadvantages such as highcost, technical complexity, labour intensity, and low sensitivity (50%) when compared to nucleic acid amplification testing (NAAT) 26 .Compared to culture, polymerase chain reaction (PCR) on cervical swab has shown very high specificity, usually ≥ 99% 21 , while a meta-analysis reported a pooled sensitivity of 88.6% 26 .
The aims of this study were to evaluate, in primary health care clinics in Manaus, Amazonas State, Brazil, the performance of the HCII CT-ID test using cervical specimens against the realtime PCR (q-PCR) assay as a gold standard; its usefulness to detect C. trachomatis cases among women under 25 years; and its operational suitability when used as an opportunistic screening tool from both the health professionals' and women's perspectives.

Study setting and population
Manaus is the capital city of Amazonas State (1,802,014 inhabitants) and has the second highest prevalence of C. trachomatis infection in Brazil 20 .The study was conducted in the Central-West Health District of the municipal public health system from the city of Manaus.This area, with two policlinics and 22 primary health care (PHC) services covering 14.1% of Manaus population, was selected because it presented a good health infrastructure system, PHC with equipped rooms for endocervical sample collection, and the health staff offered cervical cancer screening.Throughout the implementation of C. trachomatis screening, coordination between the Secretariat of Health in Manaus, the Central-West Health District, and its PHC services was established through regular meetings.
Between October 2012 and December 2013, we recruited women presenting at the two policlinics and 21 PHC services.Screening was offered during patient visits or when cytology to screen for cervical cancer was collected.We included asymptomatic women aged 14 to 25 years who agreed to sign the consent form.We excluded pregnant women and those who had used antibiotics during the previous 15 days.

Study and specimens processing
The target population were informed about the study aims during an advertising campaign that included displaying posters at health services, community health workers advertising during home visits, and educational talks at schools offered by medical students.The campaign was initiated shortly before recruitment and conducted throughout the screening program.Participants provided signed informed consent and completed a 10-minute questionnaire administered by trained health professionals who attended the services.The questionnaire included sections on sociodemographic indicators, sexual behavior, STI and HIV testing history, risk perception for STI, and substance abuse.
A nurse or doctor collected a single cervical specimen during the collection of cytology for cervical cancer screening.For C. trachomatis screening, we used a Digene cervical sampler brush (Qiagen, Mississauga, Canada) that was placed into a tube containing 1.0mL of Digene sample transport medium (STM).Samples were stored at room temperature until the end of the week, when they were transferred to the Fundação Alfredo da Matta (FUAM) laboratory and stored at -20ºC until processed.

Digene Hybrid Capture II DNA test procedure
A laboratory technician performed Digene HCII CT-ID according to the manufacturer's instructions as described previously 22 .Samples were processed in batches of up to 88 along with 8 control samples (4 positive, 4 negative).Test results were calculated by using assay-specific software accompanying the DML 2000 luminometer.Relative light units (RLUs) for the positive and negative controls were used to calculate the run-specific cutoff (CO).Specimen results were reported as RLU/CO ratios.Results were considered positive for RLU/CO values > 2.5.Samples with RLU/CO values ranging from 1.0 to 2.5 were considered equivocal and were retested.Test results were returned to the PHC.

Evaluation of the HCII CT-ID test against q-PCR
A subset of stored cervical specimens was used to evaluate the performance of HCII CT-ID against the Kit Artus CT Plus RG q-PCR 96 CE (Qiagen, Hilden, Germany) according to the assay package instructions.A laboratory technician, blind to the HCII CT results, tested all specimens that were positive by HCII CT test (n = 153) and the same number of those samples that tested negative.Negative samples were randomly selected from 15 storage boxes (ten per box) by another laboratory technician.A 300µL aliquot was transferred from each selected specimen to a 1.5mL tube before HCII CT-ID was performed.DNA was extracted using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) and the q-PCR performed on the Rotor-Gene 3000 instrument (Corbett Research, Australia).The interpretation of all q-PCR results followed the algorithms in the manufacturer's package insert.The only sample identifiers were specimen numbers.

Evaluation of operational characteristics
To identify factors that might influence test acceptability and operational performance, 52 health professionals attending the health services and all participants were interviewed.Information was gathered on service characteristics, opinions about Chlamydia screening, experience regarding STI, financial resources, time and other potential barriers to screening, and test's acceptability to participants.For participants, these questions were part of the initial interview and focused on reasons for C. trachomatis testing, STI knowledge, screening acceptability and its barriers.

Statistical analysis
Data were analyzed using Stata/SE version 11.2 (StataCorp LP, College Station, USA).Descriptive analyses were used to summarize participants' characteristics and the operational characteristics of screening to both participants and health professionals.The performance characteristics of HCII CT test compared to that of q-PCR (sensitivity, specificity, and positive and negative predictive values) were calculated by standard methods and are presented with 95% confidence intervals (95%CI).

Ethical statement
The study protocol was approved by the Ethics Research Committee of FUAM (approval number: 028/2011) and by the Ethics Committee of the Health Secretariat in Manaus.All participants signed a written consent form that was approved by the Ethics Committee.Testing with q-PCR was approved by the Ethics Research Committee of FUAM (approval number: 742.410).This study component had a waver for consent form because all samples were completely anonymized before being processed.

Subjects' characteristics
A total of 1,187 consenting women were included in this study.The characteristics of study participants by age group are shown in Table 1.Mean age was 20.1 years (SD: 2.9).Most participants (82%) had at least completed primary school, and were single or did not live with a partner (85.9%).Almost half (47.4%) were evangelical christians, 20.2% earned less than minimum wage (< US$ 230.84/month), 35.4% had had an STI, 47.8% had been tested for HIV, and 57.4% had no or low risk perception for STIs.The majority (63.3%) had their first sexual intercourse at 15 years or older, 25.5% had had more than one partner during the preceding six months, 78.7% currently had a regular partner, and 16.1% had a new partner during the preceding 6 months.Among participants, 4.9% currently exchanged money for sex, 30.1% always used a condom during sexual intercourse in the preceding 6 months.During the preceding 6 months, recreational drug use was low (0.2%), 29.5% consumed alcohol, and 57.1% practised binge drinking which was defined as having at least 4 drinks in one occasion.

Performance of HCII CT-ID test
Among the 1,187 women screened, 10 were not tested with the HCII CT-ID test and were excluded from the analysis.Eight women had an indeterminate result and were also excluded from analysis.Therefore, 1,169 participants had a HCII CT-ID test result.Among those 13.1% (153/1,169) tested positive for C. trachomatis, including 16% (82/512) of women aged 14-19 years and 10.8% (71/657) of those aged 20-25 years.Overall, 292 specimens were tested by q-PCR.Of those, one resulted in no amplification and was excluded from the evaluation of the performance analysis.The sensitivity of HCII CT-ID compared to q-PCR was 72.3% (95%CI: 65. 4

Operational characteristics
Table 2 presents reasons and potential barriers for C. trachomatis screening among young women.The mean time for patients to reach the PHC service was 15 minutes (IQR: 5-20), and most women (77.7%) reached it by walking.Only 27.9% received screening information during the advertisement campaign, mainly at the PHC service (38.6%).Only 40.6% were aware that testing for HIV and syphilis was offered at the PHC service.The main motivation for testing was that screening was offered during the campaign or at the PHC in 58.7% of the women.Overall, 80.8% women did not report problems with screening although 17.9% reported discomfort during endocervical sample collection.Screening acceptance was high (90.9%)and 98.1% were willing to test again.Overall, knowledge about STIs was high, with 92.5% of the women correctly identifying that condoms should be used before sexual intercourse starts, and 95.9% reporting that partners need treatment when someone has an STI.
The health professionals interviewed (n = 52) were 7 doctors (13.5%), 32 nurses or nurse technicians (61.5%), 8 community health agents (15.4%), and 5 social workers (9.7%).Health professionals (n = 52) found that screening for C. trachomatis is a priority (87.8%) and would support maintenance of the screening program (91.8%), although 38.8% believed that it would increase their work load.Most (76%) reported sufficient working experience to attend to STI patients, and 86% felt comfortable discussing patients' sexual health.Health professionals identified time pressure (27.4%), lack of human resources (28.5%), lack of privacy (55.8%), lack of sample collection material (26%), the existence of positive cases who did not return for their results (56.4%), lack of resources to perform active case finding of Chlamydia positive cases (53.9%), delays in delivery of testing results (46.8%), and lack of support from administrative staff (9.6%) as obstacles for offering C. trachomatis screening.Almost three-fourths (74.5%) correctly identified that screening was opportunistic (i.e. offered regardless of the women's reason for attending the PCH services), though only 45.1% reported that they would screen women with no appointment.

Discussion
This is the largest C. trachomatis screening study to date providing an evaluation of the introduction of HCII CT-ID test in PHC clinics in Brazil.The study results can help to revise the national recommendation of using HCII CT-ID, which is the only test available to detect C. trachomatis in the Brazilian public health system.The study results showed that the HCII CT-ID test had performed moderately well detecting C. trachomatis among asymptomatic women in Manaus.Although it had excellent specificity, its sensitivity was lower than that of q-PCR due in part to differences regarding the limit of detection for these assays (> 100 C. trachomatis copies for the HCII CT test compared to > 1 copy for q-PCR) 21 .Previous studies that compared the performance of HCII CT and Amplicor PCR for the diagnosis of genital C. trachomatis infection reported higher sensitivity, ranging from 93.3% to 95.4%, and specificity, ranging from 99% to 100% 22,24 .Interestingly, the HCII CT-ID-test identified a high proportion of C. trachomatis cases among young women in the 14-19 years of age group.The majority of women attending PHC services accepted the test offered as a screening tool and were willing to test again, although low STI risk perception, low awareness that screening and other HIV/STI tests were offered in PHC, and discomfort caused by sample collection were identified as barriers to engage young people in screening.In fact, most women tested simply because it was offered, suggesting that the introduction of routine screening across all PHC may encourage testing among women who would otherwise not seek it, mainly because they are asymptomatic.
From a health professionals' perspective, C. trachomatis screening was considered a priority and was offered by staff with sufficient working experience to deal with STI cases and who felt comfortable discussing sexual health.However, the fact that less than half of health professionals interviewed would test without appointment -upon which routine offer of screening to control C. trachomatis infection hinges -is cause for concern.The increase in workload reported by health professionals, which has been identified as a barrier to screening delivery, is of additional concern 17 .Other barriers identified were poor supply chain management that leads to shortage of speculum and gloves to collect material by the municipality.Given that the objective of a C. trachomatis screening program is to decrease transmission and prevent sequelaes, the success of any program is dependent not only on the ability to screen women but also on the ability to ensure treatment of those who test positive.In our study, early treatment of infection faced challenges, including infected cases who did not return for their results, delays in delivering test results to patients, and shortage of health professionals to perform active case finding.In most of the PHCs (18/22, 85.7%), service providers took advantage of routine cytological examination to offer C. trachomatis screening.Offering both screenings together might be more convenient for both the patient and the health professional.However, restricting C. trachomatis screening to only women who seek cytology would miss opportunities to test women that attend the PHC for other reasons.
This study had a few limitations.The detection of C. trachomatis by use of the HCII CT DNAtest is dependent on the number of organisms present in the specimen and may be affected by patient factors, such as the presence of symptoms 27 .Although the q-PCR assay is highly sensitive and specific 28 , it can be affected by contamination or inhibitors 29 .
In conclusion, our results suggest that HCII CT-ID is not an appropriate test to identify chlamydial organisms in endocervical samples due to its moderate sensitivity compared to that of q-PCR.Screening for Chlamydia in PHC services was feasible and well accepted by participants and health professionals who considered it a priority.In light of this finding, we recommend offering C. trachomatis routine screening for young women at PHCs in Manaus but using a test with better performance, such as q-PCR.Evidence on cost-effectiveness will be needed to guide the decision on test selection and registration in the Brazilian public health system.High levels of coverage will be needed to ensure an effective screening programme.In fact, most screening programmes have experienced difficulties in achieving the 30-50% coverage among young people needed to reduce C. trachomatis prevalence 16 .
We identified several operational barriers that may limit screening uptake.Health professionals struggled to deliver screening due to an increase in their workload and problems with supply chain management.They were reluctant to offer C. trachomatis screening without appointment and preferred to offer screening together with cytology for cervical cancer screening instead.Early treatment of positive cases faced challenges that might limit Chlamydia control efforts.Among young women, none or low risk perception of STIs, which was a risk factor for C. trachomatis infection, was also found to be a barrier to engaging them in screening.The successful implementation of screening programs for Chlamydia in similar settings will need to overcome such barriers.While still under debate, early C. trachomatis detection and treatment will produce the most benefits if opportunistic screening is offered as part of a national coordinated program.

Table 1
Sociodemographic characteristics, sexually transmitted infection (STI) and HIV testing history, STI risk perception, behavioral characteristics, and Digene Hybrid Capture II (HCII CT) test result in young women in Manaus, Amazonas State, Brazil, by age group.Total corresponds to number of patients with a HCII CT result.Of the total sample of 1,187 women, 18 were excluded from the performance analysis (10 were not tested, and 8 had an indeterminate HCII CT result).

Table 2
Reasons and potential barriers for Chlamydia trachomatis screening among 1,187 women.Manaus, Amazonas State, Brazil.