Article Text

Original article
High yield of reinfections by home-based automatic rescreening of Chlamydia positives in a large-scale register-based screening programme and determinants of repeat infections
  1. Hannelore M Götz1,
  2. Ingrid V F van den Broek2,
  3. Christian J P A Hoebe3,
  4. Elfi E H G Brouwers3,
  5. Lydia L Pars4,
  6. Johan S A Fennema5,
  7. Rik H Koekenbier5,
  8. Sander van Ravesteijn1,
  9. Eline L M Op de Coul2,
  10. Jan van Bergen4
  1. 1Department of Infectious Disease Control, Rotterdam Rijnmond Public Health Service, Rotterdam, The Netherlands
  2. 2Epidemiology & Surveillance Unit, Centre for Infectious Disease Control, National Institute of Public Health and the Environment, Bilthoven, The Netherlands
  3. 3Department of Infectious Diseases, South Limburg Public Health Service, Geleen, The Netherlands
  4. 4STI AIDS Netherlands, Amsterdam, The Netherlands
  5. 5Cluster of Infectious Diseases, Department of Research, Online Research and Prevention Unit, Amsterdam Health Service, Amsterdam, The Netherlands
  1. Correspondence to Dr Hannelore M Götz, Department of Infectious Disease Control, Rotterdam-Rijnmond Public Health Service, PO Box 70032, 3000 LP Rotterdam, The Netherlands; hm.gotz{at}


Introduction In a systematic internet-based Chlamydia Screening Implementation Programme in The Netherlands, all chlamydia-positive participants automatically received a testkit after 6 months to facilitate early detection of repeat infections. The authors describe participation in repeat testing and prevalence and determinants of repeat infection during three consecutive annual screening rounds.

Methods Data collection included information on testkits sent, samples received and results of laboratory tests at time of baseline test and retest; (sexual) behavioural variables and socio-demographic variables were assessed. Chlamydia positives were requested to answer additional questions about treatment and partner notification 10 days after checking their results.

Results Retest rate was 66.3% (2777/4191). Retest chlamydia positivity was 8.8% (242/2756) compared with a chlamydia positivity at first screening test of 4.1%. Chlamydia positivity was significantly higher in younger age groups (14.6% in 16–19 years, 8.5% and 5.5% in 20–24 and 25–29 years; p<0.01); in participants with lower education (15.2% low, 11.1% medium and 5.1% high; p<0.001) and in Surinamese/Antillean (13.1%), Turkish/Moroccan (12.9%) and Sub-Saharan African participants (18.6%; p<0.01). At baseline, 88.7% infected participants had reportedly been treated and treatment of current partner was 80.1%.

Discussion Automated retesting by sending a testkit after 6 months to all chlamydia positives achieved high retest uptake and yielded a positivity rate twice as at baseline and can therefore be recommended as an additional strategy for chlamydia control. The high rate of repeat infections among known risk groups suggests room for improvement in patient case management and in effective risk reduction counselling.

  • Chlamydia trachomatis
  • screening
  • repeat infection
  • participation
  • positivity
  • automated home-based testing
  • clinical STI care
  • surveillance
  • syndromic management
  • epidemiology
  • general practice
  • HIV
  • chlamydia infection

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Repeat infections with Chlamydia trachomatis (Ct) are common1 ,2 and remain a challenge in the control of chlamydia. Repeat infections can be caused by reinfections from an untreated sexual partner in an ongoing partnership, a new infection in the regular partner, new infections transmitted by a new sexual partner or persistent infections and treatment failure.1 ,3 With each reinfection, the chance of developing serious complications from Ct increases.4 To reduce chlamydia prevalence and incidence of complications such as pelvic inflammatory disease and subsequent infertility, screening is recommended.5 ,6 Whether a screening programme is systematic or opportunistic, treatment of diagnosed infections, preventive advice and partner notification (PN) are important parts of chlamydia control. To reduce high levels of reinfection, more attention to proper case management is advocated.7 Repeat infections may be reduced by strict regulations for PN, as reported from Sweden.8

In a previous chlamydia screening pilot in The Netherlands, the prevalence of reinfections was 10.4% (5/48) in previous chlamydia positives and 2.9% (4/139) in chlamydia negatives.9 Previous chlamydia infections are known as a risk factor for repeat chlamydia infection.10 ,11 Rescreening of infected persons within 1 year is advocated by many authors8 ,12 ,13 and recommended in national screening guidelines.6 In a systematic internet-based Chlamydia Screening Implementation Programme (CSI) in The Netherlands that started in 2008, all chlamydia-positive participants automatically received a testkit after 6 months to facilitate early detection of reinfections. We describe prevalence of repeat infection during three consecutive annual screening rounds, study participation in retesting and explore management of cases and partners at baseline.


The design of CSI has been described previously.14 ,15 In short, it is systematic, population-based, selective, for 16–29 years old sexually active persons, and has multiple screening rounds with an interval of 1 year. The screening has a regional approach, with a different selection tool for less urbanised areas. The roll-out of the screening is cluster randomised. CSI home-based testkits can be requested online after register-based postal invitation. Infected participants get a referral letter to their health provider with the advice to get treatment for themselves and their current partner; ex-partners can be alerted by the participant and request a testkit via the website. Infected participants who do not check their test result online receive it by postal letter. After 6 months, re-testkits are automatically sent to previously infected participants.

Data collection and definitions

Data collection included information on testkits sent, samples received and results of laboratory tests at time of baseline test and retest and socio-demographic variables such as age, gender and ethnicity (based on country of birth of invitee and parents). Socio-demographic variables are available for all invitees from the population registers. Participants filled in a questionnaire on background/sexual behaviour voluntarily at baseline, including educational level, history of STI, age at first sex, number of lifetime partners, number of partners during last 6 months, new sexual contact last 2 months, condom use at last contact and ethnicity of partner (see figure 1). As behavioural determinants could only be used from the questionnaire as reported at time of the first chlamydia diagnosis, analysis at time of retest was only performed for the first three variables. Chlamydia positives were requested to answer additional questions about treatment and PN 10 days after checking their results. The results of this latter questionnaire were used to describe the case management at baseline.

Figure 1

Flow of retesting and availability of data.

Assignment of community risk levels (high, medium and low) was based on the local area (cluster) profile in terms of the proportions of inhabitants from young age groups, with non-Dutch ethnicity and with low income. Socio-economic status (SES) is based on income, education level and employment, available for the lowest level postcode areas (parts of streets) (provided by The Netherlands Institute for Social Research; for more details15).

‘Participant’ is defined as a person who returned a sample to the laboratory. The participation rate was defined as the total number of participants who returned samples, divided by the number of invitees. The positivity rate was calculated as the number of positive chlamydia tests divided by the number tests performed with a valid result. Treatment percentage was defined as the proportion of test positives visiting a medical consultant and receiving treatment (self-reported) among those who responded to the treatment questionnaire. Current partner treatment percentage was assessed as the proportion of self-reported partners who were treated among all partners who were indicated by the index case as eligible for treatment.

Statistical methods

Univariate and multivariate logistic regression analyses were conducted to assess associations between participation at retesting after 6 months (dependent variable) and socio-demographic and self-reported behavioural variables (independent variables). The same analyses were conducted to study the correlates of chlamydia positivity.

Variables showing an association of p<0.20 were included in the multivariate analyses. Highly correlated variables (Spearmann correlation coefficients) were not included in the model (strongest associated variables selected). Backward stepwise logistic regression analyses were performed, including variables with a p value for the likelihood ratio test of <0.05. Results are given as ORs with 95% CIs. Adjusted ORs give estimated effects of each variable after adjustment for all other variables.

The study was approved by the Medical Ethical Committee of the Free University Amsterdam.


Rescreening after 6 months

In three screening rounds, a total of 4846 chlamydia infections were detected in 4700 individuals and 4191 (86.5%) of them were invited for retesting after 6 months. Non-invited persons either had moved out of the screening area or the 6 months period was not yet passed at the end of the project. Crude participation rate in retesting was 66.3% (2777/4191). Retest chlamydia positivity was 8.8% (242/2756) (see figure 1).

Participation at retesting

Retest participation was lower in men compared with women, lowest in the youngest age group and lower in non-Dutch participants, especially Surinamese/Antillean and Sub-Saharan African invitees. Low education level was significantly associated with low participation. Other demographic factors like screening area, community risk and SES did not significantly contribute to participation nor was there a difference between screening rounds (see table 1). In multivariate analysis, age group, education and ethnicity remained independent predictors of participation.

Table 1

Participation retesting in CSI

Since behavioural variables were only available for a small number of persons, multivariate analysis was based on socio-demographic variables only. Persons not being treated at baseline were less likely to participate than persons treated. Partner treatment was not significantly associated with participation. Persons with higher age at first sexual contact were more likely to participate in retesting than those with younger age at first sexual contact (see table 1).

Ct positivity at retesting

Ct positivity at retesting after 6 months was significantly higher in younger age groups and highest (14.6%) in the 16–19 year age group. A higher risk for repeat infection was also found in participants with lower education (15.2%). Ct positivity was 6.6% in Dutch participants but twice as high or higher in Surinamese/Antillean (13.1%), Turkish/Moroccan (12.9%) and Sub-Saharan African populations (18.6%; p<0.006). Rate of repeat infection was highest in the Rotterdam screening area versus Amsterdam and South-Limburg. Chlamydia positivity was highest in areas of high community risk and lowest in low community risk areas. There was no significant association between screening round or SES. In multivariate analysis age group, ethnicity and education remained independent predictors for repeat chlamydia infection (see table 2).

Table 2

Chlamydia positivity at retesting in CSI

The risk for repeat infection was twice as high in those participants who reported not being treated at baseline compared with those being treated (see table 2). Participants who reported that their current partner was not treated at time of baseline infection were more likely to be infected at the 6 month retest than those who reported treatment of their current partner. Having a history of STI already at the first screening test seemed to increase the risk of repeat infection, although not significantly. Younger age at first sex was associated with a higher risk of repeat infection.

Management of Ct cases at baseline

Ninety-three per cent of all participants with a chlamydia-positive result did check their result online and 7.0% were informed by letter about their result.

Of all chlamydia cases at baseline, 40.7% (1870) answered the additional questionnaire on treatment and PN completely. Of those, 8.7% (164) reported not to have visited a healthcare worker and are considered as not treated. 91.3% (1706) consulted a healthcare worker for treatment (82.5% general practitioner, 14.3% public health STI clinic and 3.2% other). Receiving treatment at this consultation was reported by 97.2% (1658/1706), leading to a treatment coverage percentage of 88.7% (1658/1870).

In univariate analysis, treatment percentages were significantly lower in men, younger age groups, lower education and non-Dutch ethnicity. In multivariate analysis age group, education and gender remained as independent predictors for treatment (see table 3 (web)). Among behavioural variables, only young age at first sex was significantly related with non-treatment.

Partner treatment management at baseline

Among the 29% chlamydia-infected participants who filled in the questionnaire about treatment of the current partner, 659/1870 (35.2%) reported that partner treatment was not applicable (no current partner). This percentage was 36.4% in women versus 31.2% in men; p<0.001.

Of the 1211 (64.8%) persons with a current partner, 970 reported the partner being treated (80.1%). Women were less likely to report treatment of current partner at the time of first infection than men (see table 4 (web)). Also younger age groups were less likely to report treatment of current partners. Like treatment of the index case, partner treatment was also less often reported in non-Dutch ethnicities with exception of persons with a Turkish/Moroccan background. Screening area and round community risk and SES were not associated with partner treatment (data not shown). In multivariate analysis, age group, education and gender remained as independent predictors for treatment of partners (see table 4 (web)).

Of the 1261 chlamydia positives with ex-partners, 76.4% reported having notified an ex-partner, 84.0% had direct contact with ex-partner and 16.0% approached ex-partners through the CSI website.


In this large-scale register-based chlamydia screening programme, automated retesting was introduced by sending a testkit after 6 months to all chlamydia-positive participants. We found an effective participation rate of 66%. This high uptake is remarkable considering the overall low and declining participation rate in our screening programme, at well below 16%. As previously reported, infected persons are more motivated to get retested.16 However, while still above 60%, retest behaviour was lower in those with low education, non-Dutch ethnic groups and younger age groups, which is in line with other studies.17 ,18 Those who reported to have been treated had a higher retest behaviour than those not being treated, suggesting more care about their own health or better motivational behaviour by their care provider.

In reports of clinic-based cohort studies with active follow-up, the percentage of retests within 6 months is between 54% and 80%.17 In regular care, improved retest rates up to 40% by sending testkits instead of a clinic visit was reported.19–21 Innovative in our study is the automated sending of testkits, integrated within the systematic programming of the screening programme, to previously infected persons. We conclude that reaching out to test positives by directly providing home-based testkits is an effective strategy to detect repeat infections.

The chlamydia positivity at retesting in this population-based cohort was 8.8%, twice as high compared with baseline (4.1%). This percentage is lower than reported in clinical studies,17 which can be explained by the nature of our, mostly asymptomatic, study population. Repeat infections can be caused by persistent infections, treatment failure, reinfections from an untreated sexual partner in an ongoing partnership or a new infection in the regular partner or a new sexual partner. Thus, suboptimal patient management might be an important reason for repeat infections. Therefore, we compared the percentage of chlamydia positives at retesting with reported treatment of index and of current partners at baseline infection. We showed that treatment of the index case at baseline as reported was 89%. This is comparable to earlier findings from opportunistic22 and systematic23 ,24 screening programmes. Of those with a current partner, 80% reported that their partner also received treatment. In a previous pilot in The Netherlands, we found confirmed partner treatment in current partners of 57%, plus an additional 18% of potential partner treatment as reported.23 Repeat infection was more often found in cases that reported not being treated at baseline (persistent infection) and in those where the current partner was not treated (reinfection).

The repeat infection rate of 11.7% in untreated participants would then suggest a clearance of almost 90%. This would be higher than clearance rates in the literature (45%–55% after 1 year).25 However in our population-based screening programme among mostly asymptomatic patients, we may have detected more existing long-term infections as we are not informed about the onset of infection. This might translate into higher clearance, although another explanation might be underreporting of treated patients.

However, despite treatment rates of 89% in participants infected at baseline testing and treatment rates of 80% in current sexual partners, we still find a rate of 5.9% of repeat infections in persons with treatment at baseline. Based on the assumption that treatment is 95% effective and the reported treatment rates are accurate, we estimate 68% to be due to newly acquired infections.

We found youngest age group, persons with low education and non-Dutch ethnicity as independent predictors for repeat infection. These are known risk factors for chlamydia in CSI,26 in previous Dutch27 as well as in international studies,11 ,12 ,28 and apparently these factors also determine participation.

Our results have several limitations as the data are sensitive to selection bias and reporting bias; Data on index and partner treatment were only available for 42% of the infected participants, and actual (partner) treatment was not verified with the treating physician in contrast to our previous study.23 Over- or underestimation of treatment rates is possible as the treatment questionnaire was less often filled in by men, non-Dutch ethnic groups, participants with lower education and those living in high-risk areas. Limited information was available at the time of retesting. More information is needed on the reasons for repeat infections and we are currently assessing biological determinants (genotyping) of reinfection. Future research should address pathogen- and host-related factors of persistence of chlamydia infections and the possibility of Azithromycine11 treatment failure.3 ,29 ,30

We advocate that retesting should be advised in all screening programmes as the percentage of repeat infections was double when compared with chlamydia positivity at baseline, thereby increasing the yield of the programme. In Ct control, we have to deal with high-risk groups like the very young and non-Dutch ethnic groups, sharing lower treatment rates, high rates of partner change and continuing risk behaviour. As Gunn10 says: ‘STI clinic clients with a recent history of gonorrhoea or chlamydia and a high risk of subsequent gonorrhoea/chlamydia may be core transmitters who could likely benefit from risk reduction, periodic screening for gonorrhoea/chlamydia, symptom recognition counselling and preventive treatment—the essential elements of STI-prevention case management’. Therefore, retesting of positives seems to be essential in improving active case finding. Retesting advice is already implemented in the US.6 Possibly retesting by automated testkits might also work for STI clinic patients.

In conclusion, automated home-based retesting of persons testing positive in our screening programme after 6 month yielded a very high uptake rate and can therefore be recommended strongly as an additional strategy for chlamydia control. Although reported (partner) treatment rates were well above 80%, the high rate of repeat infections suggests room for improvement in patient case management and in effective risk reduction counselling.

Key messages

  • In this population-based chlamydia trachomatis screening programme with automated retesting by sending a testkit, participation in previous infected persons was 66%.

  • Retest chlamydia positivity was 8.8% compared with a chlamydia positivity at first screening test of 4.1%.

  • Percentage repeat infection was highest in younger age groups, those with lower education, and in non-Dutch participants.

  • The high rate of repeat infections suggests room for improvement in patient case management.


The advisory committee: T Bakkenist, P van den Broek, T Coenen, S de Gouw, P Bindels, B Hoebee, J Land, S Morre, R van Riel, M Kivi, M van der Sande and E Steyerberg.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Funding This work was supported ZonMW, Ministry of Health, The Netherlands, grant number 124000001.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by Medical Ethical Committee Free University Amsterdam.

  • Provenance and peer review Not commissioned; externally peer reviewed.