Introduction

Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) infections are among the most common bacterial sexually transmitted infections (STIs) and disproportionately affect men who have sex with men (MSM) worldwide.1 Two large studies conducted in Thailand between 2006 and 2010 showed that MSM had approximately 30% higher prevalence of CT infection and up to five times higher prevalence of NG infection compared with men who have sex exclusively with women.2 3

CT/NG infections are associated with acquiring and transmitting HIV infection.4 In particular, rectal CT/NG infection is strongly associated with an increased risk of HIV acquisition among MSM.5 6 And while the impact of pharyngeal infection towards HIV acquisition is less understood, it is highly prevalent and may, therefore, serve as an important for infection at genital sites.7–9 Since CT/NG infections are often asymptomatic,10 the lack of routine screening may result in a missed opportunity to diagnosis these curable STIs.

The diagnosis of CT/NG infections, both at genital and extragenital sites, can be made using nucleic acid amplification tests (NAATs). Many studies have shown superior sensitivity and specificity of NAATs in detecting extragenital CT/NG infection compared with culture.11–14 The findings prompted the US Center for Disease Control and Prevention (CDC) to recommend the use of NAATs for pharyngeal and rectal CT/NG screening,15 although they have not been cleared by the US Food and Drug Administration.

Frequency of testing and anatomical sites to be tested are the two factors to consider in asymptomatic CT/NG screening. The CDC STD treatment guideline recommends that all sexually active MSM should be screened at least annually at sites of contact regardless of condom use.15 More frequent screening is advised if the individuals are at increased risk. Conversely, the Australian STI management guideline recommends screening at all sites regardless of sexually exposed contact routes.16 However, many barriers prevent the implementation of these recommendations in clinical practice. For the clients, these barriers may include the cost of tests, underestimating the risk of asymptomatic infections and concern of being stigmatised.17 18 Healthcare providers also often lack knowledge on the importance of STI screening at appropriate anatomical sites,17 which may also be the case in Thailand where there are currently no consensus recommendations for CT/NG screening.

Our primary objective was to determine the proportion of missed CT/NG diagnoses if a single anatomical site screening was performed among MSM by examining the pattern of anatomical sites of CT/NG infections. Other objectives were to determine the prevalence of CT/NG infections among MSM enrolled in the community-led test and treat cohort and to examine the prevalence of CT/NG infections in the remaining two anatomical sites if one site was negative to evaluate the proportion of missed diagnoses per individual. The findings from our study will be crucial in guiding the recommendations for CT/NG screening among MSM, both in HIV treatment and prevention programmes, in resource-limited settings.

Methods

Results

Pattern of anatomical distribution for all CT/NG infections

Of the total 349 CT infections in our study, 8.0% were isolated to pharyngeal site, 54.7% to rectal site and 22.4% to urethral site (figure 1). On the basis of our data, 85.9%, 30.6% and 67.8% of the total CT infections in our study would have been missed if only pharyngeal, rectal or urethral screening was performed, respectively. Of the total 249 NG infections, 25.3%, 37.8% and 12.5% were isolated to pharyngeal, rectal and urethral sites, respectively (figure 2). Collectively, 55.7%, 39.6% and 77.4% of NG infections would have been missed if only pharyngeal, rectal or urethral screening was performed, respectively.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Distribution of Chlamydia trachomatis (CT) infections (n=349) by anatomical site.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Distribution of Neisseria gonorrhoeae (NG) infections (n=249) by anatomical site.

Discussion

We examined the pattern of anatomical sites of CT/NG infections and showed that among MSM who tested negative for CT or NG infection at either pharyngeal, rectal or urethral site, 8%–19% had CT infection and 7%–12% had NG infection at the remaining two sites. Of the 349 CT infections, 8.0% were isolated to pharyngeal site, 54.7% to rectal site and 22.4% to urethral site; and 85.9%, 30.6% and 67.8% of the total CT infections would have been missed if only pharyngeal, rectal or urethral screening was performed, respectively. Of the 249 NG infections, 25.3%, 37.8% and 12.5% were isolated to pharyngeal, rectal and urethral sites, respectively; and 55.7%, 39.6% and 77.4% of NG infections would have been missed if only pharyngeal, rectal or urethral screening was performed, respectively. The majority of each anatomical site of CT/NG infection was isolated to their respective site, with rectal site having the highest proportion of isolation: 78.9% of rectal CT and 62.7% of rectal NG infection. These data suggest that screening at all self-report sexually exposed contact routes is highly recommended. However, if this is not feasible, screening at rectal site would provide the highest yield of CT/NG diagnoses.

The overall prevalence CT/NG infections at any anatomical sites in our cohort was comparable to the historical Thai facility-based test and treat cohort which enrolled previously unknown HIV-status Thai adult MSM and TGW with similar risk behaviours in 2012 (21.4% for CT and 12.4% for NG infection).20 The prevalence of CT/NG infections per each anatomical site in our study was comparable to one of the largest studies tested for pharyngeal, rectal and urethral CT/NG infections based on their self-reported exposure conducted in San Francisco between 2010 and 2011. Among 3039 MSM enrolled, the prevalence of pharyngeal, rectal and urethral CT infections was 2.3%, 11.9% and 4.4%, respectively; and 6.5%, 9.7%, and 5.5% for pharyngeal, rectal and urethral NG infections, respectively.21

To the best of our knowledge, our study was among the first to report the proportion of missed CT/NG diagnoses per individual if single anatomical site screening was performed. Supposing that one anatomical site screening was performed, 8%–19% of MSM who tested negative for CT infection and 7%–12% of those tested negative for NG infection actually had CT and NG infections at the remaining two anatomical sites, respectively. Importantly, the proportion of these potential missed CT/NG diagnoses increased to 11%–33% for CT and 10%–25% for NG infection among newly diagnosed HIV-positive MSM. This may be because of a higher proportion of unprotected sex and self-reported STIs in the past 6 months among HIV-positive MSM compared with HIV-negative MSM. These results point out the importance of CT/NG screening at all self-report sexually exposed contact routes. However, if resource limits the number of sites screened, rectal site proves to be the site of choice for screening, with less than 10% showed any infection in the remaining two anatomical sites if tested negative.

A study conducted in San Francisco in 2003, in which NAATs were used to test MSM for chlamydia and gonorrhoea at all three anatomical sites, was among the first published studies to show that the majority of CT (53%) and NG (64%) infections were at non-urethral sites, and would have been missed if only urethral screening was performed.10 More recent published data from multisite in USA and the Netherlands showed a range of 43%–69% of extragenital CT infection and 46%–76% of extragenital NG infection would have been missed if only urethral screening was performed, which were in line to our findings.21–24 Data from the San Francisco’s STD clinic between 2008 and 2009 showed that if one anatomical site screening was performed, screening only the pharynx would miss 81% of CT infection and 32% of NG infection; and screening only the rectum would miss 23% of CT infection and 52% of NG infection.25 Regardless of our similar findings that rectal site screening would miss the fewest infections, the high proportion of potential missed diagnoses if a single anatomical site screening was performed in any of the three sites supports the critical need for all sites, at least depending on self-reported sexually exposed contact routes, among MSM.

Although CT/NG infections at each anatomical site possess distinct characteristics, such as clinical manifestations, different duration of infections, and concerns over drug-resistant pathogens,26 27 the most important thing is the ability to detect and treat those infections regardless of site. Due to their asymptomatic nature, many patients may not be aware of the importance of the infections and do not seek medical advice.18 Healthcare provider can take the lead in encouraging sexually active MSM to screen for CT/NG infections, at least depending on their self-reported site of exposure, as the first step towards detecting and providing timely screening and treatment towards preventing transmission in the community.

Nonetheless, the cost of the test could be a major obstacle in implementing this recommendation resource-limited settings. For instance, the current cost of NAAT test for CT/NG infections in Thailand is approximately US$30 per anatomical site. This is considered expensive since more than half of our MSM participants have monthly income of less than US$320. To reduce the cost of test, a strategy to test pooled specimen has been made with promising results.28 Effort in lowering the cost of the CT/NG screening test or developing affordable molecular technologies for CT/NG detection is needed for MSM in resource-limited settings.

Certain limitations need to be considered. First, sexual behaviours were assessed using a self-administered paper questionnaire. While self-administered questionnaires may improve disclosure of sensitive behaviours, actual risk behaviours may still be under-reported. Second, risk behaviours were captured within the past 6 months. Because CT and NG infections have a long duration of infection, capturing risk behaviours within the past 6 months was beneficial in assessing risk behaviours since the potential contact date of the infections. However, the relatively long recall period may lead to recall bias. Third, because CT/NG screening in our study was based on self-reported sexually exposed contact routes rather than universal screening at all sites, we were unable to compare the performance between a history-based and universal approach. Furthermore, by limiting our analysis to MSM who self-reported sexual contact in all three anatomical sites, the findings may be biased towards those with higher risks which may have led to an overestimation of prevalence of CT/NG infections in our sample. Finally, we may have missed extragenitalia samples from a modest number of participants (248 MSM (13.4% of total MSM enrolled)) due to social desirability bias regarding questions about the site(s) of sexual contact.

Our study found that a high proportion of CT/NG infections would have been missed among MSM if single anatomical site screening is performed, especially among HIV-positive MSM. We recommend that all-site screening should be performed among MSM, at least based on self-reported sexually exposed contact routes. However, if this is not feasible, rectal screening provides the highest yield of CT/NG diagnoses. Effort in lowering the cost of the CT/NG screening test or developing affordable molecular technologies for CT/NG detection is needed for MSM in resource-limited settings.