Article Text
Abstract
Objectives We aimed to ascertain the proportion of positive, and predictive factors of chlamydia infection among females, heterosexual males and men who have sex with men (MSM) presenting to a sexual health service reporting contact with a chlamydia infected sexual partner.
Methods A cross-sectional analysis of patients attending the Melbourne Sexual Health Centre from October 2010 to September 2013. Behavioural data obtained using computer assisted self-interview were analysed to determine factors predictive of chlamydia.
Results Of the 491 female, 808 heterosexual male, and 268 MSM chlamydia contacts, the proportion diagnosed with chlamydia were 39.9% (95% CI 35.7% to 44.3%), 36.1% (95% CI 32.9% to 39.9%) and 23.5% (95% CI 18.8% to 29.0%), respectively. Female chlamydia contacts were more likely to have chlamydia if age <25 (adjusted OR (AOR) 1.86, 95% CI 1.12 to 3.10) or if they reported inconsistent condom use during vaginal sex with a regular male partner (AOR 2.5, 95% CI 1.12 to 6.14). Heterosexual male contacts were more likely to have chlamydia if age <25 (AOR 1.69, 95% CI 1.25 to 2.28) or if they had a regular female sexual partner (AOR 1.38, 95% CI 1.03 to 1.85). In MSM urethral chlamydia was diagnosed in 8.8%, rectal chlamydia in 20.2%, and 3.9% at both sites. MSM were more likely to have chlamydia if they had a regular male sexual partner (OR 2.12, 95% CI 1.18 to 3.81).
Conclusions This study of female, heterosexual male, and MSM presentations with self-reported chlamydia contact provides insight into the likelihood and predictive factors of infection. The data may inform policy and individual clinical decision making regarding presumptive treatment of chlamydia contacts.
- CHLAMYDIA INFECTION
- PARTNER NOTIFICATION
- SEXUAL HEALTH
Statistics from Altmetric.com
Introduction
Chlamydia continues to be a very common bacterial sexually transmitted disease in young men and women despite considerable investment in screening and treatment programmes.1 ,2 Estimated population prevalence in young sexually active females in the USA (2007–2012) was 4.7%,3 and 5% in young Australian females in community or general practice settings.4 While most infections with chlamydia are asymptomatic, infection has potentially serious consequences particularly for women including pelvic inflammatory disease, chronic pain, ectopic pregnancy and infertility.5
A significant proportion of people diagnosed with chlamydia will contact at least one of their sexual partners to let them know they have been exposed.6 Empiric presumptive treatment at the time of testing but prior to confirmation of infection has generally been adopted for patients who present to a clinic reporting sexual contact with a person diagnosed with chlamydia, as timely treatment is important to preventing complications and further transmission.7–9 Presumptive treatment of sexual partners also underpins the practice of partner delivered patient therapy, which is practiced widely in some countries for individuals diagnosed with chlamydia.10–13 However, as transmission is not inevitable,14 ,15 a proportion of the sexual partners will not have an infection and yet risk being prescribed antimicrobial therapy unnecessarily.
Justification for presumptive treatment for a treatable sexually transmitted infection such as chlamydia depends in part on the underlying prevalence of infection among those reporting contact with that infection. To provide these data for chlamydia, we performed a cross-sectional study to ascertain the rate of infection and factors predictive of chlamydia infection among patients presenting to a sexual health service reporting contact with a sexual partner diagnosed with chlamydia. These were compared between three groups: women, heterosexual men, and men who reported sex with men (MSM).
Methods
This study was undertaken at the Melbourne Sexual Health Centre, the main public sexually transmitted infections clinic in Victoria, Australia. In October 2010, the clinic began prospectively identifying individuals presenting to the service reporting sexual contact with a person diagnosed with chlamydia: henceforth referred to as ‘chlamydia contacts’. Chlamydia contacts were identified by the triage nurse so that these patients could be routinely offered treatment for chlamydia on the day of presentation.
During the study period chlamydia testing was undertaken using the BD ProbeTecTM strand displacement assay (Becton Dickinson, New Jersey, USA): Urethral chlamydia was tested by first void urine (FVU) or urethral swab in men, anorectal chlamydia was tested in MSM by anorectal swab. In women FVU, high vaginal swab, or cervical swabs were performed. There was no statistically significant difference between the proportion of cervical or high vaginal swabs or FVU positive for chlamydia in women, hence they were combined for our analysis. Testing was not routinely performed for pharyngeal chlamydia in either men or women, nor were anorectal samples taken in women.
Data on chlamydia contacts attending the clinic between October 2010 and September 2013 were extracted from the clinic's computer database. These data included detailed sexual behavioural data routinely obtained from clinic patients by Computer Assisted Self Interview (CASI): questions regarding numbers of sexual partners, sexual practices, and condom use. Completion of CASI was required at all first clinic visits and at subsequent visits at least 3 months apart. We excluded chlamydia contacts who did not complete CASI from further analysis.
The proportion of chlamydia contacts who were infected with chlamydia was calculated for three groups: females, heterosexual males (no male sex partners within 12 months) and MSM (males who reported sex with another male within the prior 12 months). These were compared with the overall clinic population proportions positive of chlamydia diagnosed at first clinic visits over the same period. For MSM separate analyses for rectal chlamydia and urethral chlamydia were undertaken.
Univariate logistic regression for each group was conducted to examine potential predictors of chlamydia infection among chlamydia contacts. Age was treated as a binary variable with a cut-off at 25 years for all groups. The number of sexual partners was treated as a binary variable using the median number of partners in each group as the cut-off. Crude ORs with 95% CIs were calculated. Variables with p<0.05 identified in univariate analysis were entered in the multivariate logistic analysis and adjusted ORs (AOR) with 95% CI calculated. For MSM potential predictors of infection were examined for urethral, rectal, and both anatomical sites in univariate analysis. Multivariate analysis for MSM was not performed for chlamydia as the infection could be at single or both sites. For the statistically significant predictive variables we also calculated the negative predictive values (NPV) to estimate the performance of these variables in predicting a negative chlamydia test in our population.
Where available, we examined the records of nominated sex partners (henceforth referred to as ‘partners’) of the chlamydia contacts for laboratory evidence of chlamydia infection up to 3 months prior to the presentation date of the contact.
Statistical analyses were performed using SPSS Statistics V.21. The study was approved by the Alfred Hospital Research Ethics Committee (Number 300/13).
Results
There were 2023 individuals reporting sexual contact with a person with chlamydia who presented to the clinic over the study period. Repeat presentations (N=95), those who did not complete CASI (N=316) and men who could not be identified clearly in the record as heterosexual or MSM (N=45) were excluded. In total 1567 individuals (491 females, 808 heterosexual males and 268 MSM) were included in the final analysis. Over the same period, a total of 34 924 patients attended the clinic for the first time and were tested for chlamydia.
Females
The median age among the 491 female chlamydia contacts was 24 years (IQR 22–27). The proportion of female contacts who were diagnosed with chlamydia was 39.9% (95% CI 35.7% to 44.3%). This was significantly higher than the proportion positive among female first visit attendees overall, where 5.7% (95% CI 5.4% to 6.2%) tested were positive for chlamydia (p<0.001).
Potential predictors of chlamydia infection among female contacts are shown in table 1. Factors associated with a diagnosis of chlamydia on multivariate analysis were age <25 (AOR 1.86; 95% CI 1.12 to 3.10) and inconsistent condom use with a regular male partner (AOR 2.50; 95% CI 1.12 to 6.14). Among female contacts the NPV for age <25 and inconsistent condom use with a regular male partner were 64.6% (95% CI 58.2% to 70.6%) and 75.9% (95% CI 56.5% to 89.7%) respectively. For females with both of these risk factors, 53.9% (62/115) were diagnosed with chlamydia. Among females who never used condoms with regular male partners, 51.9% (55/106) had chlamydia. There was no significant difference in the proportions positive for chlamydia according to whether women reported sex with a regular partner only (31.8%), casual partners only (37.9%) or both regular and casual partners (44.1%), p=0.287. There was no significant association between the reporting of genital symptoms and a diagnosis of chlamydia among female contacts.
Heterosexual males
The median age among the 808 heterosexual male chlamydia contacts was 26 years (IQR 23–30). The proportion of heterosexual male contacts who were diagnosed with urethral chlamydia was 36.1% (95% CI 32.9% to 39.5%). This was significantly higher than that among heterosexual male first visit attendees where 6.2% (95% CI 5.8% to 6.6%) tested positive for chlamydia (p<0.001).
Potential predictors of chlamydia infection among heterosexual male contacts are shown in table 2. Factors associated with chlamydia on multivariate analysis were: age <25 (AOR 1.69; 95% CI 1.25 to 2.28) and having a regular female sex partner (AOR 1.38; 95% CI 1.03 to 1.85). Among heterosexual male contacts the NPV for age<25 and having a regular female partner were 68.7% (95% CI 64.5% to 72.8%) and 67.1% (95% CI 62.3% to 71.7%), respectively. Of the heterosexual males who had both of these risk factors, 47.9% (69/144) were diagnosed with chlamydia. Heterosexual males who reported both regular and casual female sex partners were more likely to have chlamydia (41.1%) compared with those who had only casual partners (33%) or only a regular partner (23.5%) (p=0.025). Of the 129 heterosexual males who reported urethral symptoms 50% were diagnosed with chlamydia compared with 33% of those without symptoms (p<0.001).
Men who have sex with men
The median age among the 268 chlamydia contacts who were MSM was 30 years (IQR 25–38). Nineteen of these men were known to be HIV positive. The proportion of MSM chlamydia contacts who were diagnosed with urethral and/or rectal chlamydia was 23.5% (95% CI 18.8% to 29.0%). The proportion of MSM who tested positive for urethral chlamydia was 8.8% (95% CI 5.8% to 12.7%), rectal chlamydia was 20.2% (95% CI 15.5% to 25.7%), and 3.9% were infected at both sites. These were significantly higher than the proportions positive among MSM first visit attendees where 3.1% (95% CI 2.7% to 3.4%) and 5.5% (95% CI 5.1% to 6.1%) tested positive for urethral and rectal chlamydia, respectively (p<0.001).
Potential predictors of chlamydia infection at both sites, and specifically for rectal and urethral infection among MSM who were chlamydia contacts, are shown in table 3. Having a regular male sexual partner was associated with having urethral or rectal chlamydia (OR 2.12; 95% CI 1.18 to 3.81). Rectal chlamydia was significantly associated with receptive anal sex with casual male sex partners (OR 7.78; 95% CI 1.02 to 59.10) and within this group, with inconsistent condom use (OR 2.76; 95% CI 1.30 to 5.85). Among MSM contacts, the NPV of having a regular male partner was 83.1% (95% CI 75.5% to 89.1%), the NPV for rectal chlamydia for receptive anal sex with casual partners was 96.3% (95% CI 81.0% to 99.9%) and inconsistent condom use was 85.9% (95% CI 76.6% to 92.5%).
There were no significant associations with urethral chlamydia found. Of the 19 HIV-positive MSM, 9.5% (6/19) were diagnosed with chlamydia compared to 6.3% (13/205) of the MSM without HIV (p=0.38). Reporting of symptoms in MSM was not predictive of chlamydia infection. MSM reporting all three significant associations tested positive for chlamydia at either site in 42.9% (15/35) of cases. MSM who had both regular and casual partners were more likely to have chlamydia than those who had only casual partners (30.8% vs 16.9% p=0.01). Of the 136 MSM with regular partners, 7 did not also have casual partners, of which 3 were diagnosed with chlamydia.
Sixty-two clinical records of partners of chlamydia contacts were available for analysis. Of these partners 16/23 women, 19/25 heterosexual men and 10/14 MSM tested positive for chlamydia within 3 months of the contacts’ presentation to the same clinic. Where the partners had confirmed chlamydia, the proportion of positive chlamydia results in the contacts were 11/16 (68.8%) in females, 8/19 (42.1%) in heterosexual males and 1/10 (10%) in MSM. In females, this was significantly higher than the overall proportion diagnosed with chlamydia among contacts (68.8% vs 39.9%), p=0.021.
Discussion
In this study, a substantial proportion of individuals presenting to a sexual health service self-reporting sexual contact with a sex partner with chlamydia were found to be chlamydia infected: over a third of heterosexual males and females and just under a quarter of MSM. Within each group we identified factors that predicted a greater likelihood of chlamydia infection being diagnosed. Female chlamydia contacts were more likely to have chlamydia if they were younger or if they reported inconsistent condom use during vaginal sex with a regular male sexual partner. Heterosexual male chlamydia contacts were more likely to have chlamydia if they were younger or if they had a regular female sexual partner. MSM who were chlamydia contacts were more likely to have urethral or rectal chlamydia if they had a regular male sexual partner. In addition, MSM who reported receptive anal sex with casual male sexual partners and those who used condoms inconsistently during such encounters were more likely to have rectal chlamydia.
To our knowledge ours is one of few studies to examine chlamydia infection specifically among men reporting sexual contact with men with chlamydia. In one study validating nucleic acid amplification test detection for rectal chlamydia it was found that 39% of male partners of men with rectal chlamydia had urethral chlamydia.16 It is notable that overall in our study, chlamydia contacts who were MSM were less likely to be diagnosed with chlamydia than either females or heterosexual males. However, testing practice and performance characteristics of the assays at different sites may have underestimated true infection in this group. During the time period over which this study was undertaken the Australian guidelines did not recommend pharyngeal testing for chlamydia, hence pharyngeal chlamydia testing was not performed. Screening MSM for pharyngeal chlamydia became recommended in Australia in 2014.17–19 The performance of the BD ProbeTecTM strand displacement assay may have influenced the proportion with a positive diagnosis at different sites.20 In particular, the sensitivity of this assay for rectal chlamydia has been variably reported between less than 50% to 94.7% and so conceivably could have resulted in underdiagnoses in MSM in our study.21–23 Despite this, the higher rate of rectal compared with urethral chlamydia among MSM is consistent with other studies,24 ,25 and could in theory reflect relatively more frequent receptive anal sex, increased exposure and/or susceptibility of the rectum to chlamydia, a longer duration of rectal infection, or a combination of these.
In heterosexuals, chlamydia rates of 52–68% have been reported in contacts of sexual partners with confirmed chlamydia infection.14 ,26 ,27 In this study, the proportion diagnosed with chlamydia in the small number of contacts where chlamydia was confirmed in a sexual partner is consistent with this. Young age, inconsistent condom use, number of sexual partners, presence of symptoms, cervical ectopy, and in females an uncircumcised male sexual partner have all been described as associate risk factors for infection in contacts.14 ,26–29
Our study provides insight into the likelihood of diagnosis of chlamydia infection in the common ‘real life’ context of a clinical presentation of an individual with self-reported contact with chlamydia. We are limited, however, in what inferences we can make regarding the transmissibility of chlamydia. We did not require the confirmation of the sexual partner's chlamydia status nor specify a time period from exposure. We also did not have data on the nature or number of sexual exposures with the source partner. Nonetheless, regular sexual partners did appear to be important in the probability of chlamydia acquisition in all three groups. We speculate that this is because multiple sexual encounters are more likely to have occurred with a regular partner than would be the case with a casual partner increasing the likelihood of transmission, although these terms used to describe partners were not predefined.
Presumptive treatment of individuals who report sexual contact with chlamydia—rather than treatment based on test results would be justified by a substantial chlamydia rate among chlamydia contacts and where delayed treatment or failure of the individual to return for treatment would lead to further transmission or complications of the infection. The reliability of the performance of the diagnostic assay at the site tested may also influence whether empiric treatment is advised. The major argument against presumptive treatment is the potential for adverse drug events in individuals who are not actually infected, and also that single dose azithromycin therapy may undertreat rectal chlamydia infection.30 As a lower rate of infection was found among MSM in this study, further research using different parameters could help to inform whether presumptive treatment of all MSM chlamydia contacts is warranted or whether a more targeted approach based on risk would be preferable. In the meantime, for the three groups, our data may assist individual clinicians and their patients in their decision as to whether to treat presumptively, or to wait for test results and return to the clinic for treatment if positive.
Key messages
An expanded understanding of chlamydia infection in sexual partners may enable a targeted empirical presumptive approach to treatment of contacts.
This study examined proportions of infected and predictive factors for chlamydia in heterosexual males and females, and men who have sex with men (MSM), presenting as contacts.
Over a third of heterosexual males and females and just under a quarter of MSM were found to be chlamydia infected.
Younger age, regular sex partners, and inconsistent condom use were variably predictive of infection.
Acknowledgments
The authors would like to thank Afrizal and Jun Kit Sze for obtaining the data.
References
Footnotes
Handling editor Jackie A Cassell
Contributors SH and MC conceived and designed the study in consultation with CF and JH. All authors contributed to the analysis and interpretation of results, drafting and revisions of the manuscript and approval of the final version to be published.
Funding This research was supported by the Australian National Health and Medical Research Council (NHMRC) programme grant (No.: 568971). EPFC is supported by the Early Career Fellowships from the Australian NHMRC (No: 1091226).
Competing interests None declared.
Ethics approval Alfred Hospital Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.