Repeat Chlamydia trachomatis testing among heterosexual STI outpatient clinic visitors in the Netherlands: a longitudinal study

Chlamydia trachomatis (chlamydia) is the most frequently reported bacterial sexually transmitted infection (STI) among heterosexuals [1]. Chlamydia reinfections are also common, with proportions of reinfection of up to 32% in women [2], and 18.3% in men [3]. Chlamydia infections are associated with increased risk of complications such as pelvic inflammatory disease (PID) [4‐6]. Repeat testing for chlamydia could identify chlamydia infections at an early stage of infection, thereby possibly reducing the individual risk of complications, and might lead to reduced transmission at the population level [4‐6]. Therefore, several countries have established STI care guidelines on repeat testing for heterosexuals [7‐10].

Repeat testing guidelines differ between countries with regard to the groups that are targeted, and the advised timing of the repeat test. A mathematical modelling study estimated the risk of a repeat chlamydia infection to be highest 2–5 months after treatment, suggesting a possible window for repeat testing in case of an initial positive test [11]. This is in line with the recommended test interval in most countries for repeat testing after chlamydia infection: the US centre for disease control and prevention recommends men and women to return preferably 3 months after treatment, and at least within 12 months [7]. In Australia, repeat testing at 3 months is recommended [8], and European and UK guidelines recommend repeat testing 3–6 months after treatment for both men and women aged <25 years [9, 10]. Repeat testing after a negative chlamydia test is not explicitly recommended, but annual chlamydia screening is advised in the European and US guidelines for all sexually active men and women aged <25, or women aged <25, respectively [7, 9].

In the Netherlands, guidelines on repeat testing are limited. For men who have sex with men, advice is given to test for STI every 6 or 3 months, depending on risk behaviour [12]. For heterosexuals, guidelines on repeat testing are only established for initial chlamydia positive persons, and differ between testing facilities. In general practitioner guidelines, it is advised to retest patients within 12 months after treatment of an initial infection [13]. STI clinic guidelines recommend repeat testing 4–6 months after treatment [12]. However, while the repeat testing behaviour of MSM has been studied in the past [14, 15], it is not well known to what extend the guidelines for heterosexuals are currently being adhered to. In addition, knowledge on repeat testing among heterosexual STI clinic visitors with an initial negative chlamydia test is limited. Besides initial chlamydia positive attendees, the initial negative attendees are also of interest because Dutch STI clinics are restricted to people at increased risk of STI. Initial negative visitors could therefore still be a potential target group for repeat testing.

Two Dutch studies assessed the uptake of repeat testing after chlamydia infection at STI clinics, with differing results. One study found that 33.4% of all chlamydia positive individuals returned for a repeat test 3–12 months after treatment [16], whereas the other study found 9.2% repeat testing at 5–8 months after treatment among young heterosexuals [17]. However, these studies used regional STI clinic data only, and did not assess repeat testing behaviour and the chlamydia prevalence at repeat consultation among visitors with an initial chlamydia negative test result.

To inform further specification of the Dutch guidelines, and to be able to evaluate them properly in the future, more insight is needed into current chlamydia repeat testing behaviour among Dutch heterosexual STI clinic visitors. In this study, we aimed to describe current repeat testing behaviour and to identify groups for which repeat testing advice could be advantageous. We do this by determining repeat testing uptake and by identifying predictors of repeat testing and predictors of chlamydia positivity at repeat test among heterosexual STI clinic visitors with both an initial negative or an initial positive chlamydia test.

Overall, 15% of heterosexual women and 11% of heterosexual men visiting Dutch STI clinics between June 2014 and December 2015 returned to the same clinic for repeat testing within the study period. The strongest predictors of repeat testing were having had more than four sex partners in the past 6 months and having a chlamydia infection at the initial consultation. Reinfection rates were highest among initial chlamydia positives at three to 5 months after the initial consultation, with positivity rates over 25%. Chlamydia infection at the initial consultation and young age were strong predictors for having a chlamydia positive repeat test.

This study uses national STI clinic surveillance data, which results in a very large study population, allowing extensive analyses among both initial negative and initial positive visitors. However, there are also limitations to this study. First, linkage of consultations by ID number was only possible for individuals returning to the same STI clinic. Hence, (repeat) tests at other STI clinics or other testing facilities, such as general practitioners or online STI test providers, are not included and this might have resulted in an underestimation of the repeat testing rates. From 2016 onwards, STI clinics ask their attendees if and where they tested for STI in the past year. These new data might provide future insight in repeat testing between STI clinics and from other testing facilities to the STI clinics. Second, as this study used STI clinic surveillance data, only information on individuals who actively returned to the STI clinic was available. The positivity rates in people not returning to the clinics are unknown, which hampers generalisability of our results to all STI clinic visitors. Third, longitudinal data has only been available since June 2014, which resulted in a relative short follow-up time of 1.5 years. Consequently, we have underestimated the percentage of repeat testers as single testers might have returned for a repeat test after the follow-up period, and single testers might have been misclassified if the initial consultation in our data was actually already a repeat test. Uncertainty analyses showed that including only people with a minimum of 12 months follow-up did increase the percentage of repeat testers, but hardly changed the found determinants of repeat testing, indicating that our results were robust for follow-up time. Last, we only included the first two consultations of each individual. As only a small percentage of people had more than two consultations, we expect this exclusion not to have influenced our results.

We found that overall, 13.7% of heterosexuals visiting STI clinics returned for a repeat chlamydia test during our study period, and 20.5% when including only people with a minimum of 1 year follow-up. This is slightly lower than a study among UK genitourinary medicine clinic visitors, where a repeat chlamydia testing incidence of 26 per 100 person years was found [21]. Our study showed that 21.1% returned after an initial positive test, which was also lower than a study using US laboratory data, where repeat testing was 38% in women and 22% in men after an initial positive test [22]. Our estimate is, however, in the middle of two other regional estimates from the Netherlands (9.2% [17] and 33.4% [16] of repeat tests after an initial positive test). Differences between our study findings and others can be explained by differences in study population, definition of repeat test, and duration of follow-up. Among women, the percentage of repeat testers was higher than among men. This is in line with other Dutch studies on repeat testing [16, 17, 23]. This could be due to the fact that women are known to generally seek health-related information and health care more often than men [24, 25]. Last, we showed that repeat test positivity rates were highest among initial chlamydia positives between 3 and 5 months after the initial consultation, which is in line with findings from other studies [11, 26] and current repeat testing guidelines after CT infection elsewhere [7‐10].

Predictors of repeat testing found in our study were mostly characteristics associated with- or indicative of sexual risk behaviour and risk of STI: higher number of partners, non-Western ethnicity [18, 27], history of STI infection, and chlamydia infection at initial consultation. This might indicate that individuals who are at increased risk are also more motivated to test repeatedly. However, it could also be a consequence of the STI clinic policies, since non-Western ethnicity is a triaging criterion and therefore facilitates admittance to the STI clinic. Strikingly, we did not find young age (<25) to be predictive of repeat testing among women in our main analysis, and only a small effect among men, despite the fact that being aged <25 also is an important triaging criterion. In the uncertainty analyses with persons with a minimum of 12 months follow-up time, women aged >25 were less often repeat testers, which was more in line with the triaging criteria. However, sample sizes were much smaller in the uncertainty analysis, making these results less reliable. Future research using longer time periods is warranted to determine these associations with more certainty.

The strongest predictor for a chlamydia positive repeat test, among repeat testers, was young age. This is in agreement with studies showing that young persons are at increased risk of chlamydia infection [18, 28, 29]. Initial chlamydia infection (women) and initial chlamydia infection combined with symptoms (men) were also strong predictors for having a positive repeat test. Chlamydia infection being a predictor for repeat infection is in line with findings from studies from the UK, US, and Sweden [21, 22, 30]. Our analysis was corrected for several risk behaviour parameters at initial consultation, such as number of partners, but the association between initial and repeat infection might be due to other behavioural or biological factors as well. For example, repeat infections could be a consequence of inadequate partner notification and/or treatment of the partner of the infected index case thereby increasing the probability of reinfection of the index case [31, 32]. Furthermore, treatment failure could also contribute to the association between initial and repeat positive test. In the Netherlands, people tested positive at the urogenital site receive 1 g of azithromycin, whereas those tested positive at the anorectal site receive 7-days (100 mg twice daily) doxycycline [33]. However, treatment failure after urogenital or anorectal infections for chlamydia is possible [34, 35]. For heterosexual men, chlamydia infection was only associated with repeat test positivity if symptoms were also reported. This might indicate that for women and men different mechanisms behind repeat infections are involved. A possible explanation might be that women can be reinfected from the anorectal site to the urogenital site through autoinoculation [36]. At the STI clinics, women are only tested for anorectal infection on indication, which is likely to miss anal infections in women who did not report recent anal intercourse [37]. When women are infected at both anatomic locations, but not tested at the anorectal site, they receive azithromycin, which showed reduced effectiveness for treating anorectal chlamydia infections [35]. This may cause higher rates of repeat infections due to autoinoculation in women, whereas for heterosexual men, autoinoculation is less likely [38].

When comparing the most important predictors of repeat testing with predictors of having a chlamydia positive repeat test among repeat testers, several important findings emerge. First, initial chlamydia infection was associated with both repeat testing and with repeat test positivity. This is an indication that initial positives are indeed at increased risk of repeat infection. It also shows that chlamydia positives are already returning to the STI clinic more often compared to initial chlamydia negatives, following guidelines. However, still only 21% of initial positives were repeat tester. Second, young age (<25) was not associated with repeat testing in the main analysis, and was a strong predictor of a chlamydia infection at repeat test. This indicates that young people may not be returning to the STI clinic more often, even though they are at the highest risk of infection at repeat test compared to the other ages. Therefore, young heterosexual STI clinic visitors could be an important target group for enhancing repeat testing. Third, non-Western ethnicity was associated with repeat testing. However, at repeat test, no difference in chlamydia positivity was seen between persons with non-Western ethnicity compared to persons with Dutch ethnicity. This finding was unexpected, as more STIs are generally found among those with non-Western ethnicity [18, 27]. This might indicate that the selection of persons with non-Western ethnicity that do return to the STI clinics for a repeat test, are not necessarily the ones that are at highest risk of obtaining STIs. Or, vice versa, that those with a Dutch ethnicity returning for a repeated test are the ones with increased risk of STI. The same thing was seen for number of partners; having more partners was associated with repeat testing, but not with positivity at the repeated test. This could also indicate that those returning are not necessarily the ones at highest risk.

Altogether, the results of this study provide an extensive overview of current STI repeat testing behaviour among Dutch heterosexual STI clinic attendees, which can be used as guidance to determine for which groups repeat testing could be most advantageous. Our findings suggest that repeat testing could especially be beneficial for young persons (<25 years) with an initial chlamydia infection. However, interpretation of the results remains challenging due to the use of surveillance data, and since the motivations of people to return (or not return) to the clinic remain unknown. Furthermore, there is little insight into the current practices of STI clinics regarding whether (because of long waiting lists and high workload) and to whom repeat testing advice is given. Gaining insight into current practices is important to identify barriers and attitudes toward repeat testing among health care professionals, which should be taken into account when evaluating the guidelines. More research is needed to determine repeat testing strategies in the Netherlands in terms of (cost) effectiveness, feasibility, and public health impact.