Diagnosis and Management of Uncomplicated Chlamydia trachomatis Infections in Adolescents and Adults: Summary of Evidence Reviewed for the 2021 Centers for Disease Control and Prevention Sexually Transmitted Infections Treatment Guidelines

Abstract

To prepare for the development of the 2021 Centers for Disease Control and Prevention (CDC) sexually transmitted infections treatment guidelines, the CDC convened a committee of expert consultants in June 2019 to discuss recent abstracts and published literature on the epidemiology, diagnosis, and management of sexually transmitted infections.This paper summarizes the key questions, evidence, and recommendations for the diagnosis and management of uncomplicated Chlamydia trachomatis (CT) infections in adolescents and adults that were reviewed and discussed for consideration in developing the guidelines. The evidence reviewed mostly focused on efficacy of doxycycline and azithromycin for urogenital, rectal, and oropharyngeal CT infection, CT risk factors in women, performance of CT nucleic acid amplification tests on self-collected meatal specimens in men, and performance of newer CT point-of-care tests.

Chlamydia trachomatis (CT) infection remains the most prevalent reportable bacterial sexually transmitted infection (STI) in the United States (US), with >1.8 million cases reported to the Centers for Disease Control and Prevention (CDC) in 2019 [1]. Young age remains a strong predictor of CT infection, with the highest CT infection rate in persons aged <25 years [1]. With most CT infections being asymptomatic, screening for CT is necessary to detect most infections. The 2015 CDC sexually transmitted diseases (STD) treatment guidelines recommend annual CT screening in all sexually active women <25 years of age and in older women with risk factors (eg, women with a new sexual partner or multiple partners, a partner with concurrent partners, or a partner with an STI) [2]. Although CT screening efforts in the US may have contributed to a decline in rates of pelvic inflammatory disease (PID) [3, 4], which may translate to lower rates of PID complications including infertility, the number of reported CT infections in the US is at an all-time high; this suggests we need to reconsider approaches to the diagnosis, management, and prevention of CT infection in order to decrease CT infection rates.

Since the release of the 2015 CDC STD treatment guidelines, there have still been questions and knowledge gaps in the epidemiology, diagnosis, and management of uncomplicated CT infection that needed to be addressed. To prepare for the development of the 2021 CDC STI treatment guidelines, the CDC convened a committee of expert consultants in June 2019 to discuss recent abstracts and published literature on the epidemiology, diagnosis, and management of STIs. This article summarizes the key questions, evidence, and recommendations for the diagnosis and management of uncomplicated CT infections in adolescents and adults that were reviewed and discussed for consideration in developing the guidelines. The review focused on study findings that became available after the evidence review that was conducted for the 2015 CDC STD treatment guidelines [5].

SUBJECTS AND METHODS

A search of the literature from 15 February 2013 through 11 February 2019 was conducted using the PubMed/Medline computerized database of the US National Library of Medicine and was limited to studies involving humans as subjects. The search included the Medical Subject Heading (MeSH) terms “Chlamydia Infections” and “Chlamydia” and the keywords “chlamydia” and “chlamydial” in titles and abstracts (tiab). This search yielded 3433 citations. A second search using the same time period was conducted to identify publications that had not yet been indexed in Medline. The search included the terms “chlamydia” and “chlamydial” and was limited to publications in English. To limit the number of duplicate citations from the first search and to focus on the newest papers, especially those not indexed for Medline, the words “NOT medline[sb]” were added to the search terms. The second search yielded 1072 citations.

Papers solely discussing lymphogranuloma venereum, trachoma, or Chlamydia species other than CT were excluded from our review, as were those not including adolescents or adults. Papers were then selected and reviewed for new or updated data that could address key questions (listed below in the Results) on the epidemiology, diagnosis, or management of CT infection. Papers that addressed the Key Questions were then summarized in an evidence table that outlined their study design, methodology, results, and conclusions; the table of evidence is available online at: https://www.cdc.gov/std/treatment-guidelines/evidence.htm. The evidence and the discussions with the expert consultant committee were then used to address the key questions and to develop suggested recommendations for the 2021 CDC STI treatment guidelines.

RESULTS

Key Question 1: Are New Data Available on Risk Factors for CT Infection in Women That Could Help to Guide CT Screening Recommendations?

In the 2015 CDC STD treatment guidelines, CT infection screening in women was recommended annually for sexually active women<25 years of age (the age range in which CT prevalence is highest) and in older women with risk factors, including sex with a new partner, multiple partners, a partner with a concurrent partner(s), or a partner with an STI [2]. Papers in the current literature review that specifically evaluated risk factors for incident or prevalent CT infection were reviewed.

There were 5 longitudinal studies identified that evaluated risk factors (biological and/or behavioral) for incident CT infection [6–10]:

• Hwang et al evaluated CT nucleic acid amplification test (NAAT) results and behavioral and biological data collected from a human papillomavirus virus (HPV) study cohort of women 13–22 years of age recruited from a family planning clinic or college health center in San Francisco (from October 2000 to April 2012) [6]. Women were seen every 4 months for behavioral interviews and specimen collection and cervical CT NAAT was performed annually (and at interval visits where symptoms or possible STI exposures were reported). Analyses for this study included 629 women who attended at least 1 follow-up visit. The median number of visits was 17 and follow-up time 6.9 years. They found that 15% of women had incident CT infection, and risk factors identified in the final multivariate model were HPV infection at the preceding visit, smoking, and weekly use of substances besides alcohol and marijuana since the prior visit.

• Aghaizu et al evaluated risk factors for incident CT infection in 954 female students aged 16–27 years recruited from 20 London universities and further education colleges for the Prevention of Pelvic Infection trial who returned self-collected vaginal swabs by mail 11–32 months (median, 16 months) after recruitment for CT NAAT [7]. Among the 907 women who tested CT negative at their baseline study visit, the estimated annual CT incidence rate was 3.4 per 100 person-years. Risk factors for incident infection (after adjusting for age) were a new sexual partner during 12 months of follow-up, smoking, concurrent bacterial vaginosis, and concomitant infection with high-risk carcinogenic HPV strains.

• Matson et al evaluated feelings about relationships and incident CT and Neisseria gonorrhoeae (NG) infection risk in a cohort of 122 sexually active adolescent women who were enrolled in clinics and community venues in Baltimore from December 2009 to August 2010 for the Perceived Risk for Sexually Transmitted Diseases study [8]. The women completed a computer-assisted self-interview, had urine CT NAAT done every 3 months, and completed daily surveys on a smartphone for up to 18 months. They found that for each week there was a decrease in partner trust, there was a 45% increase in risk for being infected with CT and/or NG (relative risk [RR], 1.45 [95% confidence interval {CI}, 1.18–1.78]; P =.004); neither a decrease in partner closeness or commitment, nor an increase in perceptions in partner concurrency or risk for STI was associated with incident CT and/or NG infection. Risks associated with just CT infection were not reported.

• Harder et al evaluated incident CT infection in 7998 women in Copenhagen previously enrolled between May 1991 through January 1993 in an HPV study who had an interview on sexual and health behaviors, cervical CT NAAT, and HPV DNA testing by Hybrid Capture 2 test at baseline, and then the same study procedures were repeated at a 2-year follow-up visit [9]. Incident CT infection occurred in 1.4% and was associated with a higher sexual partner number (odds ratio [OR], 1.07 per partner [95% CI, 1.02–1.11]), low education level (OR, 1.69 [95% CI, 1.11–2.56]), and high-risk HPV positivity (OR, 1.66 [95% CI, 1.06–2.58]); older age and condom use were associated with a decreased risk for incident CT infection.

• Russell et al evaluated incident CT infection in 225 women aged 15–35 years with genital infections or at risk for CT infection who were enrolled at an STD clinic or at a hospital-associated ambulatory clinic or research clinic in Pittsburgh from February 2011 through May 2014 [10]. At enrollment, a medical and sexual history was obtained and a cervical swab was collected for CT NAAT. Subjects returned for follow-up visits at 1, 4, 8, and 12 months for repeat medical and sexual history and repeat collection of a cervical swab for CT NAAT. There were 149 (66.2%) women with CT infection at enrollment and 48 (21.3%) with incident CT infection (defined as a positive CT test at follow-up, which for some patients with CT infection at baseline could reflect repeat infection or persisting infection). Incident risk factors for CT infection in an adjusted analysis were cervical CT infection at enrollment, an NG infection diagnosis, and sexual exposure to an uncircumcised partner (in the prior 3 months) or a CT-infected partner during follow-up.

There were also 2 cross-sectional studies identified that evaluated risk factors for prevalent CT infection [11, 12]:

• One study retrospectively performed CT NAAT on urine samples from a subset of 1136 females aged 15–17 years who participated in a population-based survey in Germany (the German Health Interview and Examination Survey for Children and Adolescents) from 2003 to 2006 [11]. The primary objective of this retrospective study was to evaluate CT infection prevalence and associated risk factors. The CT prevalence was 2.2% (95% CI, 1.4%–3.5%). In a multivariate model, 2 risk factors were identified: marijuana consumption often or several times in the preceding 12 months and general health status less than “very good.”

• In the other study, Jørgensen et al evaluated the association of concurrent partnerships and gap length between partners with CT prevalence in Danes 15–29 years of age who either had documented current CT infection (including 130 woman) or were from a randomly selected group of individuals from a general Danish population (including 1707 women) with unknown CT infection status [12]. In women, an association with CT prevalence was demonstrated for concurrent partnerships (OR, 8.5 [95% CI, 4.6–15.5]) and a shorter gap between serial monogamous relationships (OR,7.1 [95% CI, 3.6–13.8]).

In summary, there were 7 new studies identified in the current literature review that specifically evaluated risk factors for incident or prevalent CT infection in women [6–12]. The studies were heterogenous in the risk factors that were evaluated as well as the study populations. More than 1 of the studies identified smoking and HPV infection as risk factors [6, 7, 9]. Risk factors identified in only 1 of the above studies included concurrent partnerships, shorter gap lengths between serial monogamous relationships, decreasing trust in one’s partner, use of marijuana, use of substances other than alcohol or marijuana, a bacterial vaginosis diagnosis, an NG infection diagnosis, and sexual exposure to an uncircumcised male partner or partner with chlamydia [6–12]. Concomitant bacterial vaginosis had also been identified as a CT risk factor in the previous literature review for the 2015 CDC STD treatment guidelines [5].

Quality of evidence: Moderate

Strength of recommendation: B

Recommendation: Additional potential CT risk factors that may prompt CT screening in women not already undergoing screening include HPV infection and bacterial vaginosis, and these risk factors should be highlighted in the 2021 CDC STI treatment guidelines.

Key Question 2: What Is the Prevalence of Rectal CT Infection in Women and Associated Risk Factors, and How Often Does Rectal CT Infection Occur in Women in the Absence of Concomitant Urogenital CT Infection?

There were 3 systematic reviews [13–15] and an additional 9 studies not included in the reviews [16–24] that reported data on rectal CT infection in women. All 3 reviews varied in the number of articles reviewed based on their search terms, dates, and inclusion and/or exclusion criteria.

• The most inclusive review, by Chan et al, included 33 studies listed in PubMed through 1 December 2015, and reported rectal CT prevalence data in women [13]. Their main findings were that (1) rectal CT prevalence ranged from 2.0% to 77.3%; (2) most studies were done on patient populations in STD clinics or high-STI risk settings; (3) most infections were asymptomatic (36%–100% of infections); (4) a “significant” number of infected women did not report anal sex; (5) extragenital CT screening (pharyngeal and rectal) increased the yield of detecting CT over only doing urogenital CT screening by 6%–50%, and (6) risk factors for rectal CT infection were heterogenous across studies but included younger age, partners with chlamydia, anonymous partners, a partner who is an injection drug user, trading sex for money, and having sex under the influence of alcohol or drugs. Another finding worth mentioning was that the median rectal CT prevalence in women was similar to that in men who have sex with men (MSM) (8.7% vs 8.9%, respectively).

• A review by Dukers-Muijrers et al included 23 articles identified in a Medline search through 8 June 2015, with data on rectal CT prevalence in women [14]. Their main findings were that (1) rectal CT prevalence data had only been studied in clinic-based populations; (2) most studies included women reporting receptive anal intercourse; (3) rectal CT prevalence was higher in most studies in Canada (11.7%–13.5%) and the US (5.1%–27.3%) compared with studies in Europe (5.6%–12.5%) (excluding 2 studies that did rectal CT testing based on known genital CT infection); (4) between 5% and 29% of rectal CT infections occurred in the absence of genital CT infection; and (5) in women with urogenital CT infection, rectal CT infection was detected in 33%–83%.

• The most restrictive review, from Chandra et al, focused on rectal CT prevalence in heterosexual women in high-income countries [15]. Several databases were searched through 2 September 2015, and their review included only 14 published studies with rectal CT prevalence data on women, excluding conference abstracts and case reports. Their main findings were that (1) 5 of the studies only tested women for rectal CT if they reported receptive anal intercourse; (2) all studies were in women attending sexual health settings; (3) rectal CT positivity ranged from 1.7% to 77.3% across the studies, and therefore due to the heterogeneity across studies, a summary estimate was calculated for select studies; (4) summary rectal CT positivity for women attending clinics for routine care was 6.0% (95% CI, 3.2%–8.9%); (5) summary rectal CT positivity for women reporting receptive anal intercourse was 25.9% (95% CI, 8.5%–43.3%); (6) summary concurrent rectal CT positivity in women testing positive for urogenital CT was 68.1% (95% CI, 56.6%–79.6%); (7) the percentage of women with rectal CT positivity who had a negative urogenital CT test ranged from 0% to 11.5%, with considerable heterogeneity across studies; (8) summary rectal CT positivity in women attending clinics for routine care who tested negative for urogenital CT was 2.2% (95% CI, 0%–5.2%); (9) excluding studies only including women who reported receptive anal intercourse, the frequency of reported anal intercourse was 13.5%–50% in rectal CT–positive women; and (10) reported receptive anal intercourse was not associated with rectal CT positivity (summary risk ratio, 0.9 [95% CI, .75–1.10]).

Of the 9 additional studies not included in the reviews, 4 studies only evaluated women reporting receptive anal intercourse [16–19]. In those 4 studies, all women were tested at STI/sexual health clinics, the sample size of women having rectal CT NAAT performed ranged from 13 to 2878, and the rectal CT prevalence ranged from 8.2% to 15%. All 4 studies evaluated the proportion of rectal CT infections that may not have been adequately treated if only urogenital CT testing had been performed (ie, because the urogenital CT NAAT was negative) and the proportions ranged from 0% to 23% [16–19]; 1 of the studies only had 2 women with rectal CT infection [17], and if that study was excluded, then up to 18%–23% of rectal CT infections may not have been adequately treated if only urogenital CT testing were done. In the remaining 5 studies that evaluated rectal CT positivity in women and did not restrict to include only populations reporting receptive anal intercourse, the sample size of women having rectal CT NAAT performed ranged from 60 to 5499, and the rectal CT infection prevalence ranged from 5.3% to 13.4% [20–24]; 1 of the studies only included human immunodeficiency virus (HIV)–infected women [20], and therefore if this study were excluded, the rectal CT infection prevalence would range from 8.8% to 13.4%. Four of the 5 studies evaluated the proportion of rectal CT infections that would have been missed if only urogenital CT testing had been performed and the proportions ranged from 0% to 34.7% [21–24]; the 0% came from a study with a sample size of only 60 (with only 5 rectal CT infections [21], and if that study were excluded, then the percentage of rectal CT infections missed would range from 18.8% to 34.7%. Among the 5 studies not restricted to women reporting receptive anal sex, 4 evaluated 1 or more risk factors for rectal CT infection [20–23]. The Travassos et al study of rectal CT prevalence in HIV-infected women in Salvador, Brazil, reported that rectal CT and/or NG infection was associated with age<29 years and being pregnant, but not with reported anal intercourse; risk factors for just rectal CT infections (without NG infection) were not reported [20]. The Deiss et al study in women in the US Navy seen for an annual gynecology visit in San Diego reported that a younger median age was associated with rectal CT and/or NG infection; however, they did not report factors associated with just CT infection [21]. The van Liere et al study in women visiting an STI clinic in the Netherlands reported that younger age was the only determinant of rectal CT positivity (highest risk in those age≤21 years), and having an indication for rectal CT testing (self-reported anal sex or anal symptoms) was not associated with rectal CT positivity; 72.4% of women had rectal CT testing without an indication [22]. The Tao et al study evaluating data from a large US commercial laboratory reported that younger age (15–24 vs 25–60 years) was associated with rectal CT positivity [23].

In summary, studies that evaluated rectal CT prevalence in women in clinic-based settings reported prevalence estimates ranging between 5% and 15% [13–24]. None of the studies evaluated rectal CT infection in women in a general population. Several of the studies evaluated the proportion of rectal CT infections that may have been missed by only performing urogenital CT testing and found that 1 in 10 to 1 in 3 women would have had a rectal CT infection missed with only urogenital CT testing [13–16, 18, 19, 21–24]; however, the studies did not evaluate the clinical significance of rectal-only infection, including the development of rectal symptoms or subsequent urogenital CT infection. Limited studies evaluated risk factors for rectal CT infection in women and found that younger age was a risk factor, while reported receptive anal sex was not associated with rectal CT infection [13, 15, 20, 22].

Quality of evidence: Moderate

Strength of recommendation: B

Recommendation: For women who are already receiving CT treatment with an antimicrobial regimen considered effective against rectal CT (see Key Question 6 below), there is no strong justification for performing rectal CT testing. For women who have receptive anal intercourse and rectal symptoms, rectal CT testing should be performed. Anorectal CT screening can be considered in asymptomatic women who have receptive anal intercourse as well as in asymptomatic women who do not report receptive anal intercourse but are being evaluated in high-CT-prevalence settings. There is insufficient evidence to recommend rectal CT screening in women from a general population or low-CT-prevalence venue who do not report receptive anal intercourse.

Key Question 3: How Well Does CT NAAT Perform on Self-Collected Meatal Swabs in Men and What Is the Acceptability of Self-Collecting This Specimen?

Self-collection of meatal specimens for CT NAAT in men could further facilitate CT screening efforts, especially if more CT infections are detected by these self-collected specimens compared to first-catch urine (FCU). In the previous literature review [5], there had been sparse studies on the performance of NAAT on self-collected meatal swabs and findings were inconsistent; therefore, more studies were needed before a recommendation could be made regarding whether self-collected meatal specimens should be recommended as an acceptable specimen for urogenital CT NAAT screening in men.

There were 4 studies identified in the current literature review that evaluated CT detection in self-collected meatal swabs by CT NAAT and compared performance to CT NAAT on FCU and/or a clinician-collected urethral swab [25–28]:

• Chernesky et al evaluated CT NAAT on meatal Aptima swabs that were self-collected under supervision and FCU in 356 men enrolled at a healthcare center for street youth in Canada; data on frequency of symptoms were not reported [25]. They found a higher CT detection frequency in the meatal swab vs FCU (40/354 [11.3%] vs 33/356 [9.3%]; P = .039).

• Dize et al evaluated CT NAAT on a self-collected meatal flocked swab vs clinician-collected urethral swab in 203 men seen at an STI clinic in Baltimore in whom the clinician-collected urethral swab was being collected as part of routine care [26]. Subjects were≥17 years of age, 94% were black, and 63% had symptoms. The self-collected meatal swab was always collected after the clinician-collected urethral swab and subjects were given instructions on the procedure. CT-discordant results between the 2 swab specimens were tested with an alternative NAAT. CT was detected in 32 subjects (15.8%) and the meatal swab detected CT in 1 more subject than the urethral swab. With the clinician-collected urethral swab as the reference, the CT NAAT sensitivity on the meatal swab was 96.8% and did not significantly differ from the urethral swab. Most men (90.1%) rated the self-collection as very easy or easy.

• Ferrero et al evaluated CT NAAT on a self-collected meatal swab vs clinician-collected urethral swab vs FCU in 284 men seen at an STI clinic [27]. The self-collected meatal swab was collected first (following verbal instructions and a picture guide), followed by FCU, then a clinician-collected urethral swab. Sensitivity and specificity of the self-collected meatal swab were based on a CT infection status being considered true when either the FCU or clinician-collected swab had the same CT NAAT result as the self-collected swab. Most men were symptomatic (94.4%). CT was detected in 34 (11.97%) men. The sensitivity and specificity of the self-collected meatal swab was 91.1% and 99.2%, with an overall agreement of 97.7% with clinician-collected urethral swab and 90.4% with FCU. Collecting the clinician-collected swab after FCU could have lowered its CT detection rate.

• Berry and Stanley evaluated CT NAAT on a self-collected meatal swab vs FCU in 1517 males seen at a sexual health clinic in Coventry, United Kingdom, for STI screening; data on frequency of symptoms were not reported [28]. Subjects were given instructions on the self-collected meatal swab, which was collected before FCU. CT discordant samples were tested by an alternative NAAT. CT was detected in 160 (10.5%) subjects. CT concordance between the samples was 98% (1497/1517). After the alternative NAAT was performed on discordant samples, it was found that FCU detected CT in 11 more subjects. CT NAAT on a meatal swab had a sensitivity and specificity of 92.0% and 99.6%, respectively.

In summary, there were 4 studies in this literature review on the performance of CT NAAT on a self-collected meatal swab vs clinician-collected urethral swab and/or FCU, and findings suggest that CT NAAT has a comparable performance on self-collected meatal swabs as with the other urogenital specimens [25–28]. The 1 study that evaluated acceptability of self-collected meatal swabs reported that >90% of men found it to be easy or very easy [26].

Quality of evidence: Moderate

Strength of recommendation: B

Recommendation: There is moderate evidence that performance of CT NAAT on self-collected meatal swabs is comparable to that of clinician-collected urethral swabs and FCU and limited evidence that this specimen collection strategy is acceptable. Use of this strategy would require simple, focused instructions and may be most suitable for males wanting to self-collect a urogenital specimen for CT screening and who are either unable to provide urine or prefer to self-collect a meatal swab over providing urine. Self-collected meatal swabs should not be used for performing urethral Gram stains to guide clinical management as there is insufficient evidence on the use of self-collected meatal swabs for this purpose.

Key Question 4: Are New Data Available on the Performance of CT Point-of-Care (POC) Tests Compared With Conventional (Non-POC) CT NAAT?

There were several new studies, including a systematic review, published on CT POC test performance. In the systematic review by Herbst de Cortina et al, they performed a PubMed search per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines from January 2010 to August 2015 for articles on POC tests for CT (and NG and Trichomonas vaginalis) [29]. They identified 8 reports that evaluated 10 CT POC tests and 2 reports that studied POC tests for both CT and NG; all studies used NAAT as a reference. The key findings were as follows: (1) the Cepheid GeneXpert CT/NG test (NAAT) performed the best among the POC tests evaluated, with sensitivities and specificities>97% across different sample types (this test is US Food and Drug Administration [FDA] approved for urogenital CT/NG testing in the US); (2) the aQcare Chlamydia TRF kit (a lateral flow immunoassay) and the Automated Urine Flow Cytometry test performed reasonably well, with sensitivities and specificities >90% (however, they were each only evaluated in 1 study); and (3) the other POC tests evaluated had poor performance, with sensitivities and/or specificities<75%. Included in the review were also 2 studies that evaluated cost-effectiveness of CT POC tests using mathematical models, and both studies found that POC testing (using a highly sensitive test) was a cost-effective strategy [30, 31].

There were several studies evaluating performance or benefits of the GeneXpert CT/NG test that were not included in the Herbst de Cortina et al review [29]. Three studies evaluated its performance on urogenital specimens [32–34]:

• One study was performed in an urban emergency department setting in the US and found that the kappa agreement between the GeneXpert and transcription-mediated amplification (TMA) for CT was 0.81 (95% CI, .59–1.00) for men and 0.85 (95% CI, .80–.90) for women [32].

• Another study was done on women and men seen in 12 primary healthcare settings in Australia and found that the concordance between the GeneXpert CT/NG test and the conventional NAAT used in each health center was 99.4% (95% CI, 99.1%–99.7%) [33].

• A third study was done on randomly selected self-collected vaginal swab specimens from 50 HIV-infected pregnant women seen at 3 primary healthcare facilities in South Africa [34]. The GeneXpert CT/NG test (done on the first specimen) was compared to polymerase chain reaction (PCR), with an alternative PCR used when results were discordant. CT was detected in 26 (52%). Concordance between the assays was 96%; 2 samples were positive by the GeneXpert assay but negative by both PCR assays.

Three studies evaluated the performance of the GeneXpert CT/NG test on rectal swabs [35–37]:

• One study evaluated the performance of GeneXpert vs PCR on 396 self-collected rectal swabs from MSM, transgender women, and female sex workers in Papua New Guinea and found that the overall agreement between the assays was 96% for 326 samples with valid results; 70 samples had invalid results, suggesting inhibition [35].

• A second study evaluated the performance of GeneXpert vs TMA (used as the reference) on 448 self-collected rectal swabs from men and women participating in an internet-based STI screening program (with sample collection at home) in the US (www.iwantthekit.org) [36]. The sensitivity and specificity of the GeneXpert were 95.5% (95% CI, 77.2%–99.9%) and 99.7% (95% CI, 98.5%–100%), respectively, for 372 specimens with valid results; 68 samples had invalid results, mostly due to absence of human DNA in the sample.

• A third study evaluated the performance of GeneXpert vs TMA on 396 self-collected rectal swabs from men seen for routine STI testing at an AIDS wellness clinic in Hollywood [37]. Only 3 swab specimens gave invalid results by the GeneXpert assay. The concordance between the assays was 97.7% (95% CI, 95.7%–99.0%).

One study by Jacobsson et al evaluated the analytical sensitivity and specificity of the GeneXpert assay for CT as part of a World Health Organization laboratory validation of the assay [38]. They spiked 339 samples with CT strains that were phenotypically and/or genotypically diverse and with other non-CT Chlamydia species. Two different TMA assays were used as the reference standard. They found that all CT-positive samples were detected at≤10² genome equivalent concentration per reaction using the GeneXpert assay and no false-positive specimens were detected in the samples with non-CT Chlamydia species strains.

Four studies evaluated benefits and/or cost-effectiveness of the GeneXpert CT/NG testing used in routine clinical care [39–42]; however, benefits and cost-effectiveness were not separated for CT vs NG. The 4 studies are summarized as follows:

• Rivard et al conducted a retrospective quasi-experimental study in which they studied the impact of implementing the GeneXpert for rapid diagnostic testing for urogenital CT/NG in an emergency department (ED) at an urban community teaching hospital in Michigan [39]. They evaluated test results and outcomes for 200 patients tested during a time period in which the GeneXpert was used (December 2014–January 2015) compared to the same number of patients tested by traditional NAAT during an earlier time period in the same calendar months (December 2013–January 2014). Exclusion criteria included patients who left before being assessed for need for empiric therapy and those diagnosed with pelvic inflammatory disease. Subjects in the GeneXpert group were more likely to receive appropriate treatment during their ED visit (72.5% vs 60%; P = .008) and were notified with their test results more quickly (median, 17.4 vs 51.5 hours; P = .01). An estimated US$37 000 (based on the institution’s own pricing for drugs, ED visits, etc) would be saved annually using the GeneXpert.

• Wingrove et al retrospectively evaluated the benefits of GeneXpert by comparing outcomes of asymptomatic patients screened with the GeneXpert at outreach or specialized clinics in London vs asymptomatic patients seen at a sexual health clinic in London whose specimens were tested by conventional NAAT offsite [40]. Patients testing positive for CT and/or NG by GeneXpert (28 of 196 [14%]; 4 CT, 20 NG, 4 both CT/NG; 93% were MSM) were matched to those testing positive by the conventional NAAT based on type of infection diagnosed and date of sample collection. Similarly, the first 50 patients testing negative by GeneXpert were matched to the first 50 testing negative by conventional NAAT. Type of specimens tested were not specified in the publication. In patients testing CT and/or NG positive, those in the GeneXpert test group were informed of their results a median of 7 days quicker (1 vs 8 days) and treated a median of 8 days quicker (2 vs 10 days). For patients testing negative, those in the GeneXpert group were informed of their results a median of 11 days quicker (1 vs 12 days).

• Bourgeois-Nicolaos et al reported that in a time period (July 2012–July 2013) when they implemented genital CT/NG testing with the GeneXpert assay for women being seen at 2 family planning clinics in France for preabortion consultation visits (48 hours before the abortion procedure), they detected CT and/or NG in 53 of 330 women (16%), with a mean time to results of 2.5 hours; all infected women were treated before their procedure [41]. In comparison, when using conventional NAAT, they had found that 40% of infected women were not treated before their procedure.

• May et al conducted a pilot randomized controlled trial (RCT) comparing testing and treatment outcomes based on CT and NG testing using the GeneXpert CT/NG POC test (the “test group”) vs a standard-of-care conventional PCR assay (the “control group”) in 70 male and female patients presenting with urogenital symptoms to an urban ED in Washington, D.C.[42]. Randomization was intended to be 1:1, although 42 were in the case group and 28 in the control group. Both subjects and providers were informed of the testing strategy after randomization. In subjects testing negative for CT and NG, the control group had a higher proportion empirically treated for infection (ie, “overtreated”) vs the test group (11/20 [55%] vs 8/37 [21.6%]; P = .01). Based on a follow-up phone interview, test group subjects were more likely to be notified of their test results than control group subjects (difference of 50.6%) and less likely to report missing antibiotic doses (difference of 51.3%).

There were several studies evaluating POC tests other than the GeneXpert CT/NG test that were not included in the Herbst de Cortina et al review [43–50]:

• Two studies evaluated rapid CT tests that had poor performance [43, 44]. One study reported that the Chlamydia Rapid Test had a sensitivity of 20% compared to PCR on urine specimens from 100 asymptomatic males in the general population in Durban, South Africa [43]. The other study reported that the Cortez Onestep Chlamydia Rapicard Insta Test had a sensitivity of 71.4% compared to strand displacement amplification on endocervical swab specimens from 242 pregnant women attending their first antenatal clinic visit in Zimbabwe [44].

• Three studies evaluated performance of POC tests for CT that used loop-mediated isothermal amplification (LAMP) and could be performed in <1 hour [45–47]. Two of the studies reported high sensitivity (>91%) and specificity (>95%) for the LAMP-based assays when used for testing endocervical specimens [45, 46]. The third study used a LAMP-based assay that was combined with antimicrobial peptide lysis and was found to have a sensitivity of up to 73% and specificity of 100% on male and female urine specimens [47].

• Four studies evaluated performance of POC tests with other amplification technologies that performed well and had performance times ranging from 15 to 90 minutes [48–51]. Two studies evaluated the io single module system (Atlas Genetics Ltd) vs a conventional commercially available NAAT on self-collected vulvovaginal swab specimens and included an additional NAAT for testing when initial results were discordant: Harding-Esch et al found the assay to have a sensitivity and specificity of 96.1% and 97.7%, respectively [48], and Widdice et al reported a sensitivity and specificity of 92.9% and 98.8%, respectively [49]. A study evaluating a recombinase polymerase amplification-based assay (TwistDx) reported a sensitivity ranging from 94.3% to 100% and specificity ranging from 99.7% to 100% on male and female urogenital specimens [50]. A study evaluating a cross-priming amplification assay reported a sensitivity and specificity of 98.2% and 100%, respectively, on genital swab samples in women [51].

In summary, the Cepheid GeneXpert CT/NG POC test performs well against comparator NAATs across different specimen types and its use in clinical settings could expedite time to patients receiving test results and treatment. The 90-minute completion time for the GeneXpert may make it a feasible CT POC test for clinical settings such as EDs where patients may normally wait for extended periods of time to be evaluated; even outside of those clinical settings, use of the assay is likely to expedite treatment compared to conventional non-POC CT NAATs. There are some other promising POC tests in development that performed well but are not currently FDA approved in the US. There are other select POC tests that have poor performance for CT detection and should not be used.

Quality of evidence: Strong for the GeneXpert assay/ Moderate for other POC tests

Strength of recommendation: B

Recommendation: At least 1 FDA-approved CT POC test that has excellent performance for CT detection is now available, the Cepheid GeneXpert, and its use may expedite time to results notification and treatment for patients. Other promising POC tests based on NAAT are still in development and have good performance and even shorter times to test completion than the GeneXpert assay.

Key Question 5: Is There Any New Evidence on the Association of Azithromycin With Serious Adverse Events or Deaths in Those Who Receive Azithromycin for Treatment of STIs?

There has been concern that because of the possibility that azithromycin may prolong the QT interval, this could lead to a potentially fatal arrhythmia in those who may be at increased risk (eg, those already with existing QT interval prolongation, on medications that prolong the QT interval, with specific electrolyte abnormalities). A single retrospective study evaluated persons with CT and/or NG infections who received azithromycin and whether they had mortality within 10 days after treatment (from cardiovascular or other causes) [52]. Case report data came from the Oregon Public Health Division (from 1996 to 2012) and Public Health–Seattle and King County (from 1993 to 2010) and was matched to death-record data. There were no cardiovascular deaths among the 162385 persons receiving azithromycin; there were 3 deaths that were not due to cardiovascular causes.

Quality of evidence: Strong

Strength of recommendation: A

Recommendation: New data evaluating mortality risk support the safety of azithromycin for treatment of CT and NG infections.

Key Question 6: What New Data Are Available on the Efficacy of Doxycycline and Azithromycin Regimens for Treatment of Rectal CT Infection?

One meta-analysis of observational studies [53] and 6 additional observational studies [54–59] were found that addressed efficacy (ie, microbiological cure) of azithromycin and/or doxycycline for rectal CT infection. The meta-analysis by Kong et al identified 8 observational studies that evaluated the efficacy of these antibiotics for rectal CT infection [53], including an observational (retrospective cohort) study by Khosropour that was also found in the current literature review [54]. The random-effects pooled efficacy for azithromycin (from the 8 observational studies) was 82.9% (95% CI, 76.0%–89.8%) and for doxycycline (from 5 observational studies) was 99.6% (95% CI, 98.6%–100%); the efficacy difference was 19.9% (95% CI, 11.4%–28.3%) in favor of doxycycline [53]. There was considerable heterogeneity across the studies; all studies except 1 had follow-up times of >3 weeks and 2 studies tested at timepoints after 3 months.

The other 5 observational studies include the following:

• A recently published prospective observational study (Femcure) by Dukers-Muijrers et al evaluated the effectiveness of azithromycin and doxycycline for urogenital and rectal CT infections in women with a recent positive CT NAAT on a vaginal and/or rectal swab [55]. The women returned to 1 of 3 public health STI clinics in the Netherlands for treatment (at a median of 8 days after screening). All women with a positive rectal CT NAAT (including those with a positive vaginal CT NAAT) were treated with doxycycline per routine CT treatment, whereas those with a positive vaginal CT NAAT and either a negative rectal CT NAAT or a rectal CT NAAT not performed ended up receiving azithromycin for routine CT treatment. Women were enrolled at this treatment visit at the clinic in which they self-collected vaginal and rectal swabs, and then they collected swabs at home weeks 1 and 2 followed by self-collection of swabs at the clinic at week 4; swabs collected at the 4-week clinic visit were tested by NAAT for microbiological cure. There were 341 women with rectal CT infection at enrollment, of whom 319 (93.5%) had concomitant urogenital CT infection. The rectal CT infection microbiological cure rate was significantly lower for women receiving azithromycin vs doxycycline (164/209 [78.5%] vs 126/132 [95.5%]; difference, 17% [95% CI, 9.6%–24.7%]; P< .001); subanalyses in a restricted subset of women (in whom certain confounders had been addressed) showed the cure frequency ranged from 82% to 91% for azithromycin and 97% to 100% for doxycycline, with a difference of 9% to 15% between the treatments in favor of doxycycline.

• A retrospective review of rectal CT infection treatment efficacy by Gratrix et al evaluated 476 patients (219 women [46%] and 257 MSM [54%]) managed at 1 of 2 STI clinics in Alberta who returned for a test of cure (TOC) by CT NAAT 20–60 days after treatment [56]. They found no significant difference in the frequency of microbiological cure in those who received azithromycin vs doxycycline (421/460 [91.5%] vs 0/16 [100%]; P =.63); the azithromycin treatment failure rate was comparable in women vs MSM (7.1% vs 9.7%). The small sample size of doxycycline-treated subjects was a major limitation.

• Hathorn et al conducted a retrospective review of the efficacy of doxycycline for rectal CT infection in 532 patients attending a sexual health clinic in Birmingham, United Kingdom, who returned for a TOC; the proportion of patients who were male vs female was not provided [57]. Azithromycin had been the first-line treatment used in the clinic for rectal CT infection until it had been changed to doxycycline in October 2010, and this study evaluated rectal CT treatment outcomes from that date through May 2014. After reclassifying 21 patients in their TOC analysis who may have been misclassified as treatment failures (because of reinfection risk and/or nonadherence to treatment), they found the doxycycline cure rate was 99.1% (527/532). The study did not provide information on patient gender or timing of TOC.

• A retrospective review of rectal CT infection treatment efficacy by Li et al evaluated 526 patients (142 [27%] women and 384 [73%] men [99% that were MSM]) seen at the Adelaide STI clinic in Australia [58]. Rectal testing had been performed only on those reporting receptive anal sex. Prevalence of rectal CT infection was 6.7% in men and 8.1% in women; the majority (63%) were asymptomatic. Efficacy analyses were limited to patients with repeat rectal CT NAAT performed 14–180 days after treatment, which included 173 persons treated with doxycycline and 31 treated with azithromycin. A repeat positive rectal CT test occurred less frequently in those treated with doxycycline vs azithromycin (5.8% vs 19.4%; P =.01). Of note, the doxycycline regimen duration was 10 days. After adjusting for demographic, clinical, and behavioral factors, they found that azithromycin-treated patients had a significantly higher risk of having a repeat positive rectal CT NAAT (adjusted RR, 2.96 [95% CI, 1.16–7.57]).

• A retrospective study by Kong et al evaluated azithromycin efficacy for rectal CT infection in 203 asymptomatic MSM with rectal CT infection who attended the Melbourne Sexual Health Centre and were treated with azithromycin [59]. Analyses were limited to MSM with repeat rectal CT NAAT within 100 days of treatment. After excluding 32 subjects classified as reinfection, they found the azithromycin treatment efficacy was 83.6% (143/171; 95% CI, 77.2%–88.8%).

At the time of the literature review, there had been no rectal CT infection RCTs published. However, there were 2 rectal CT infection RCTs published after the literature review whose findings were considered in the CT treatment recommendations for the 2021 CDC STI treatment guidelines [60, 61]. Lau et al conducted a double-blind, placebo-controlled RCT of azithromycin vs doxycycline for asymptomatic rectal CT infection in MSM seen at 5 sexual health clinics in Australia [60]. The primary outcome was microbiologic TOC at 4 weeks posttreatment by NAAT. In the modified intention-to-treat study population (who were randomized, had a TOC NAAT, did not have lymphogranuloma venereum [LGV], had concordant CT OmpA genotypes, and did not have unprotected sexual reexposures prior to the TOC), a microbiologic cure occurred in 281 of 290 (96.9% [95% CI, 94.9%–98.9%]) in the doxycycline group vs 227 of 297 (76.4% [95% CI, 73.8%–79.1%]) in the azithromycin group, demonstrating that doxycycline was superior to azithromycin (adjusted risk difference, 19.9% [95% CI, 14.6%–25.3%]; P< .001). Dombrowski et al conducted a double-blind, placebo-controlled RCT of azithromycin vs doxycycline for rectal CT infection in MSM (who could have rectal symptoms, but not proctitis) seen at sexual health clinics in Seattle and Boston [61]. The primary outcome was microbiologic TOC at 4 weeks posttreatment by NAAT. In the complete case study population (who were randomized, CT NAAT positive at enrollment, and had a TOC NAAT), a microbiologic cure occurred in 70 of 70 (100% [95% CI, 90%–100%]) in the doxycycline group vs 48 of 65 (74% [95% CI, 56%–86%]) in the azithromycin group, showing that doxycycline was superior to azithromycin (absolute difference, 26% [95% CI, 16%–36%]; P<.001). Of note, there were 4 MSM in each treatment group infected with an LGV strain, of whom all 4 treated with doxycycline (the 7-day regimen) were cured vs 3 of 4 treated with azithromycin being cured.

There have been 3 studies published that addressed biological factors that may contribute to rectal CT infection treatment failure from azithromycin:

• Kong et al evaluated the pharmacokinetics of a single dose of azithromycin 1g in rectal tissue in 10 healthy males recruited from the University of Melbourne [62]. Men self-collected rectal swabs prior to taking azithromycin, then 2 hours and 24 hours posttreatment, and then on days 2, 3, 4, 7, 10, and 14; the men also provided data on sexual behaviors and drug side effects after treatment. Azithromycin levels in rectal swabs were measured by liquid chromatography–mass spectrometry and were found to peak at a median of 24 hours and to stay above the minimum inhibitory concentration (MIC) of azithromycin for CT for 14 days in all men. Data collected in the study suggested that increased gastric pH could increase drug levels and diarrhea and rectal douching with water-based lubricants could decrease levels.

• In the above retrospective rectal CT infection treatment study by Kong et al[59], CT organism load was measured in stored rectal swabs in 227 MSM with baseline rectal CT infection, of whom 64 (28.2%) had repeat rectal CT detection by NAAT within 100 days of treatment. Baseline median CT organism load was higher in those with vs without subsequent repeat CT detection (4.4 vs 3.8 log₁₀ copies/mL; P =.03). In a subanalysis, the baseline median CT organism load was higher in 29 MSM classified as treatment failure vs MSM without a repeat positive rectal CT NAAT (5.2 vs 3.5 log₁₀ copies/mL; P< .01).

• In an in vitro study of the susceptibility of 20 CT strains (from the most prevalent CT serovars [D, E, F, G]) to macrolides (azithromycin and erythromycin), doxycycline, and levofloxacin in colorectal and endocervical cells (Caco and HeLa cells, respectively), Foschi et al found that macrolides had higher MICs and minimal bactericidal concentrations in the colorectal cells vs the endocervical cells, whereas these values did not differ between cell lines for doxycycline or levofloxacin [63]. Azithromycin MICs in colorectal cells were higher with increasing numbers of CT elementary bodies (organism load).

In summary, there is now substantial evidence, both from observational studies and RCTs, that doxycycline is a more efficacious treatment for rectal CT infection than azithromycin and that azithromycin’s efficacy is consistently below 95% [53–61]. There were also sparse studies identified in this literature review on biological factors that may influence azithromycin’s efficacy for rectal CT infection [59, 62, 63].

Quality of evidence: Strong

Strength of recommendation: A

Recommendation: The 7-day doxycycline regimen is the recommended treatment regimen for rectal CT infection in men and women while the azithromycin 1g regimen is an alternative regimen because of its lower efficacy. When nonadherence to the multidose doxycycline regimen is a significant concern in patients being treated for rectal CT infection, then the azithromycin 1g regimen can be considered as an alternative treatment option.

Key Question 7: What New Data Are Available on the Efficacy of Doxycycline and Azithromycin Regimens for Treatment of Uncomplicated Urogenital CT Infection?

There were several publications, including reviews and additional individual studies, published since the last literature review [5] that addressed the efficacy of azithromycin and/or doxycycline regimens for uncomplicated urogenital CT infection. The following publications included studies that were conducted in clinical settings in which repeat CT exposure and nonadherence could be a confounding factor:

• A meta-analysis of 23 urogenital CT infection RCTs of azithromycin vs doxycycline was conducted by Kong et al that presented data on microbiological cure within 3 months of treatment [64]. They reported a pooled efficacy difference in favor of doxycycline of 1.5%–2.6%; however, the fixed effects pooled efficacy for azithromycin was 96.2%. A subanalysis showed a greater pooled efficacy difference in favor of doxycycline of 7.4% for symptomatic men. The majority of studies (17 [74%]) used a CT test less sensitive than NAAT (ie, culture, enzyme immunoassay, or direct fluorescent antibody) and there was considerable heterogeneity between studies for azithromycin efficacy.

• Khosropour et al conducted an evaluation of the impact of doxycycline adherence on STI outcomes, including CT infection microbiological failure, from a dataset from a previous NGU RCT by Manhart et al[65, 66]. Imperfect doxycycline adherence was defined as missing≥1 dose during the 7-day treatment regimen. Of 184 men who had been randomized to doxycycline, 28% reported imperfect adherence [66]. Microbiological failure occurred in 3 of 47 (6.4%) men with chlamydia. Among men with chlamydia infection, imperfect doxycycline adherence was associated with a 9.3-fold higher frequency of microbiological failure (95% CI, 1–89.2).

• Not included in the Kong meta-analysis [64] was a study by Kissinger et al that reexamined azithromycin efficacy data for CT urethritis among heterosexual men that had been derived from 3 earlier prospective studies; the purpose was to try to harmonize methodologies across the studies to limit inaccuracies in classification of the microbiological outcome, either from a false-positive TOC (from testing too early posttherapy) or reinfection (from reexposure) [67]. They found that the azithromycin treatment failure rate among 242 men with chlamydia varied from 6.2% to 12.8%, depending on analysis adjustments or exclusions.

• In a recently published retrospective data review of 234733 women in Washington state with urogenital CT infection treated with azithromycin (n = 168301) or doxycycline (n = 66432) between 1992 and 2015 and who had repeat CT testing performed 14–180 days after treatment, repeat CT was detected in 6.1% [68]. The frequency of repeat CT detection was higher in those who received azithromycin vs doxycycline (6.7% vs 4.7%; P < .001). An adjusted analysis (including demographics) showed that women receiving azithromycin had a 1.24 higher repeat CT detection rate (95% CI, 1.19–1.30), which did not significantly differ in subanalyses based on retesting window.

• In the recently published prospective observational study (Femcure) by Dukers-Muijrers et al that was described above, the efficacy of azithromycin and doxycycline for urogenital CT infection was evaluated in 394 women with urogenital CT infection at enrollment, of whom 319 (81%) had concomitant rectal CT infection [55]. The urogenital CT infection cure rate did not significantly differ for women receiving azithromycin vs doxycycline (246/263 [93.5%] vs 125/131 [95.4%]; difference, 1.9%, P =.504); subanalyses in a restricted subset of women (in which certain confounders had been addressed) showed that the cure frequency ranged from 94% to 99% for azithromycin and 96% to 100% for doxycycline. The study did not evaluate for repeat CT detection beyond 4 weeks.

Limitations of prior urogenital CT infection treatment studies evaluating azithromycin and/or doxycycline included the inability to control for reexposure to CT in study participants and the potential for medication nonadherence. Two studies identified in the literature review had a study design that addressed these 2 limitations:

• Geisler et al conducted a urogenital CT infection RCT of azithromycin vs doxycycline from December 2009 through April 2014 in males and females residing in youth correctional facilities in Los Angeles [69]. Of 567 participants enrolled and randomized, there were 155 participants in each treatment arm comprising the per protocol population: 65% male and 35% female. There were no treatment failures (based on urine CT NAAT at day 28) in the doxycycline arm (95% CI, 0%–2.4%). In the azithromycin arm, treatment failure occurred in 5 participants (3.2% [95% CI, .4%–7.4%]); treatment failure was higher in males (4 [3.9%]) than in females (1 [1.9%]). The failure rate difference in the treatments was 3.2%, with a 90% upper CI of 5.9%, exceeding the predetermined 5% cutoff for establishing azithromycin noninferiority. A subanalysis of the 102 males who received azithromycin showed that the treatment failure rate was higher in males reporting painful urination vs those without this symptom (2 of 20 [10.0%] vs 2 of 82 [2.4%]; P =.172); the study was not powered to evaluate the difference in azithromycin efficacy in males based on urogenital symptom status.

• Beyda et al conducted a prospective study of azithromycin efficacy for urogenital CT infection in adolescents in a youth correctional facility in Texas between May 2012 and July 2013, with a 3-week posttherapy TOC performed by urine NAAT at an average of 29 days (range, 14–136 days) after azithromycin treatment [70]. Among 128 CT-infected adolescents (99 males, 29 females), 5 subjects (3.9% [95% CI, 1.2%–8.9%]) had treatment failure; the efficacy was higher in females (100% [95% CI, 88%–100%]) than in males (95% [95% CI, 88.6%–98.3%]). Those with treatment failure more often had symptoms at follow-up (60% vs 1.3%; P < .01).

In January 2019, a Cochrane Systematic Review was published on urogenital CT infection RCTs in males and females up until June 2018 that evaluated CDC-recommended treatment regimens for urogenital CT infection: doxycycline vs azithromycin and doxycycline vs ofloxacin [71]. The review used a fixed-effects meta-analysis model and estimated the pooled risk ratio for microbiological failure. Of 14 selected RCTs, mostly in STI clinic populations, they found that the risk ratio for failure in men was higher with azithromycin vs doxycycline (RR, 2.45 [95% CI, 1.36–4.41]), the failure risk ratio for azithromycin vs doxycycline in women was uncertain (RR, 1.17 [95% CI, .48–6.16]), and the risk ratio for failure for doxycycline vs ofloxacin in men and women was uncertain; there were only 2 RCTs in men and 1 RCT in women that studied ofloxacin.

In summary, since publication of the 2015 CDC STD treatment guidelines [2], there has been 1 urogenital CT infection RCT of doxycycline vs azithromycin [69], other non-RCT urogenital CT infection treatment studies evaluating doxycycline and/or azithromycin [55, 65–68, 70], and 2 meta-analyses of urogenital CT infection RCTs published [64, 71]. Overall, the current evidence suggests that doxycycline and azithromycin regimens are both effective for urogenital CT infection in women, whereas in men, the microbiological cure rate may be a little lower for azithromycin, especially in males with urogenital symptoms.

Quality of evidence: Strong

Strength of recommendation: A

Recommendation: Although studies support that azithromycin still has high efficacy for urogenital CT infection in women, there is concern about the effectiveness of azithromycin for concomitant rectal CT infection, which studies suggest can occur in a substantial proportion of women with urogenital CT infection and is not associated with reported anal receptive intercourse (discussed in Key Questions 2 and 6). Inadequately treated concomitant rectal CT infection in women who have urogenital CT infection could increase the risk for continued CT transmission and possibly put women at risk for repeat urogenital infection (reinoculation of the urogenital site from the rectum) and its complications, although untreated rectal CT infection leading to repeat urogenital infection has not been sufficiently studied. In men, azithromycin still has a moderate to high efficacy for urogenital CT infection, although efficacy may be influenced by the presence of urogenital symptoms. Concomitant rectal CT infection is also a concern in men with urogenital CT infection, which raises the same concerns about azithromycin being an inadequate treatment for concomitant rectal CT infection in men. It would be impractical and costly to perform rectal CT testing in all men and women being treated for urogenital CT infection. Based on the above rationale, the 7-day doxycycline regimen is the recommended treatment regimen for urogenital CT infection in men and women, while the azithromycin 1g regimen is an alternative regimen for urogenital CT infection. When nonadherence to the multidose doxycycline regimen is a significant concern in patients being treated for urogenital CT infection (such as adolescents), then the azithromycin 1g regimen is an alternative treatment option.

Key Question 8: Is There Any New Evidence on the Efficacy of Doxycycline or Azithromycin Regimens for Treatment of Oropharyngeal CT Infection?

There was a single new study published that evaluated efficacy of these CDC-recommended regimens for oropharyngeal (OP) CT infection. It was a 13-month prospective open-label observational study that recruited from a hospital-affiliated clinic in Birmingham, United Kingdom, in which all male and female patients with OP CT infection diagnosed by TMA were offered azithromycin for the initial 7 months and then all patients were offered doxycycline for the remaining 6 months [72]. Of 142 subjects evaluated in the final analysis, treatment failure at 4–6 weeks posttreatment was higher in the 78 subjects treated with azithromycin vs the 64 taking doxycycline (10% vs 2%; P =.04); all of the azithromycin treatment failures were in women and the single doxycycline treatment failure was in an MSM. They excluded patients having oral sex posttreatment, having vomiting or diarrhea after taking azithromycin, and those not completing the doxycycline regimen within the anticipated 7-day period.

In summary, a new observational study suggests that doxycycline may be more efficacious than azithromycin for OP CT infection [72]. There has not been an OP CT RCT comparing these 2 treatment regimens published to date. More studies are needed that evaluate the effectiveness of different treatments for OP CT infection.

Quality of evidence: Low

Strength of recommendation: C

Recommendation: Considering that this single study suggests doxycycline may be more efficacious than azithromycin for OP CT, doxycycline is the recommended treatment regimen for OP CT infection and azithromycin can be an alternative treatment regimen when treatment adherence with doxycycline is a concern.

Key Question 9: Are There Any New Treatment Studies on Antimicrobial Agents for CT Infection Other Than the Antimicrobials Currently Recommended?

There were only 2 new published studies that addressed new antimicrobials for the treatment of urogenital CT infection, each evaluating an antimicrobial agent not FDA approved in the US [73, 74]. Takahashi et al prospectively evaluated sitafloxacin 100mg twice daily for 7 days for nongonococcal urethritis (NGU) in heterosexual men seen in clinics in Japan [73]. Of 47 men with CT NGU detected by urine CT NAAT, microbiological cure by NAAT was reported in 45 (95.7%) at a TOC visit 2–6 weeks after treatment. There was no comparator drug and it is unclear how many subjects had repeat CT testing at≤21 days, whose results could be confounded by a false-positive test result due to residual CT nucleic acids. Sitafloxacin is not currently available in the US. Geisler et al performed a double-blinded, multicenter RCT of oral single-dose rifalazil 25mg (a new rifamycin with potent anti-CT in vitro activity) vs single-dose azithromycin 1g for urogenital CT infection in women enrolled at 5 US study sites [74]. The TOC visit was at study day 22–25. Microbiological cure based on CT NAAT on a vaginal swab was lower in the rifalazil group (n = 33) vs the azithromycin group (n = 38) (85% vs 92%), with a cure rate difference of −7.3% (95% CI, −22.5% to 7.9%). Rifalazil was well tolerated.

In summary, there were 2 new studies identified that studied CT treatment other than those currently recommended. One study demonstrated high efficacy of a 7-day sitafloxacin course for CT NGU [73]. Another study comparing rifalazil to azithromycin for urogenital CT infection in women found that rifalazil may have a lower efficacy than azithromycin, although the sample size of the study was small [74].

Quality of evidence: Low

Strength of recommendation: C

Recommendation: Considering sitafloxacin’s lack of availability in the US and the single study showing lower urogenital CT infection cure rates with rifalazil (vs azithromycin), neither of these treatment regimens would offer any advantage to the recommended regimens.

Key Question 10: Are There New Data on Timing of CT Nucleic Acid Clearance After Treatment That Will Be Useful for Determining Timing of Repeat CT Testing Using NAAT After CT Treatment?

In the 2015 CDC STD treatment guidelines, TOC following CT treatment was recommended for pregnant women with chlamydia by NAAT at 3–4 weeks after completing therapy [2]. In addition to its impact on routine clinical care, TOC timing by NAAT is also highly relevant to CT infection treatment trials. A concern with testing too early following treatment is the possibility of a false-positive CT NAAT due to detection of residual CT nucleic acids from dead organisms, while a concern for delaying testing too long after treatment is the risk for repeat CT exposure from a partner with chlamydia leading to a positive TOC that may represent reinfection rather than treatment failure. Since the last literature review [5], there have been 4 studies published that have evaluated timing of clearance of CT nucleic acids after treatment [75–78]:

• Lazenby et al evaluated CT ribosomal RNA (rRNA) clearance (by TMA) in 47 women (including pregnant and nonpregnant) seen at an obstetrics and gynecology clinic in the southeastern US who completed a protocol that included a baseline self-collected vaginal swab prior to azithromycin treatment and then weekly vaginal swabs self-collected at clinic for 4 weeks [75]. They found that the median number of days till a negative CT rRNA result was slightly longer in pregnant vs nonpregnant women (median, 8 days [range, 4–29 days] vs 7 days [range, 5–13 days], respectively; P = .04).

• Versteeg et al evaluated CT clearance in endocervical swabs by CT culture and RNA- and DNA-based CT NAATs in 90 asymptomatic women seen in an STI clinic in Amsterdam who were cervical CT RNA positive [76]. Endocervical swabs were collected prior to azithromycin treatment and then at 7-, 21-, and 49-days posttreatment. At baseline, 81 (90%) women were CT DNA positive and 69 (76.7%) culture positive. Regarding the CT RNA–positive women, 78 completed follow-up and 66 (84.6%) cleared RNA by 21 days posttreatment. Regarding the CT DNA–positive women, 70 completed follow-up and 61 (87.1%) cleared DNA by 21 days posttreatment. Regarding culture-positive women, 58 had complete follow-up data; 3 (5.2%) were CT culture positive at 7 days posttreatment and none were culture positive at 21 days posttreatment. There were 5 women with CT NAAT–positive results at 49 days posttreatment and 3 were culture positive; all 5 reported unprotected sexual contact, making it difficult to rule out reinfection.

• Williams et al evaluated CT DNA clearance in 123 adolescent females who participated in a longitudinal STI study in Indianapolis and were treated with azithromycin (by prescription or directly observed therapy) [77]. Women had a cervical or vaginal swab collected at the time of treatment and then self-collected vaginal swabs weekly for up to 12 weeks after treatment. The median time to a first CT DNA–negative test was 9 days (range, 0–84 days). There were 88 (71.5%) women who were CT DNA negative by day 14 and 109 (88.6%) by day 21. Among a subset of 59 women who received azithromycin as directly observed therapy, 74.6% were CT DNA negative by day 14 and 91.5% by day 21. The authors noted that continued detection of CT DNA (presumably beyond day 21) could represent continued DNA exposure (from live or dead CT organisms) or possible treatment failure.

• Dukers-Muijrers et al evaluated duration of CT positivity after treatment by testing CT rRNA (by TMA) and CT load (by quantitative PCR) in 52 CT-infected patients seen at an STI clinic in South Limburg in the Netherlands [78]. The 52 patients had 59 CT infections: 44 cervicovaginal CT infections and 15 anorectal CT infections (7 in women and 8 in MSM). Subjects self-collected swabs at the infected site for 18 time points from the time of pretreatment through day 51. The baseline CT load was higher in anorectal vs cervicovaginal CT infections. The first CT-negative sample occurred within 9 days for 46 (78%) of the infection cases. The range of time to the first CT negative sample was 2–21 days; 2 infections never had a CT-negative sample. Among the 35 infections in patients classified as low reinfection risk (reported no sex, had sex with a treated partner, or had safe sex with a new partner), 40% had CT rRNA detection in 1 or more samples 3–8 weeks posttreatment. Intermittent CT detection occurred more often in anorectal vs cervicovaginal CT infections (2.1 times higher number of CT detections in anorectal cases).

In summary, there were 4 studies in this literature review that provided data on duration of CT RNA and/or DNA positivity after treatment with azithromycin [75–78]. All 4 included urogenital CT infections and 1 included anorectal CT infections [78]. Only 1 study included pregnant women [75]. The studies were heterogenous in terms of retesting interval, tests used, and patient populations. The majority of patients tested negative for CT DNA or RNA by 3 weeks after treatment with azithromycin. Three of the studies conducted CT testing beyond 1 month after treatment [76–78], making it difficult to exclude CT reexposure. In the single study that evaluated CT rRNA clearance in pregnant women weekly for 4 weeks, all women were CT rRNA negative by day 29 [75]. Overall, the heterogeneity in the studies on CT nucleic acid clearance posttherapy to date prohibits a more precise estimate of the time to CT nucleic acid clearance after treatment.

Quality of evidence: Moderate

Strength of recommendation: B

Recommendation: Given that some persons with chlamydia infection may still be CT DNA or RNA positive at 3 weeks after completing azithromycin therapy, it is reasonable to perform TOC in pregnant women with chlamydia at 4 weeks after completing therapy.

Key Question 11: Is There Any New Evidence to Further Support That Expedited Partner Therapy for CT Infection Can Lower Rates of Repeat CT Detection After Treatment?

There were 2 studies identified in the current literature review that evaluated whether expedited partner therapy (EPT) was associated with lower repeat CT detection frequency after treatment [79, 80]:

• Taylor et al reviewed medical records of 492 chlamydia cases (83% women) and EPT uptake from January 2009 to August 2011 from an urban Indian Health Service facility in Phoenix, Arizona [79]. They found that of 472 chlamydia cases that had documented treatment, 52% received EPT (with azithromycin), more often in women and in cases seen in a women’s clinic or primary care clinic, in asymptomatic patients, and in patients treated from 1 to 14 days after diagnosis. Of 324 patients with retesting after treatment, repeat CT detection was less common among those receiving EPT (13% vs 27%; OR, 0.5 [95% CI, .3–.9]).

• Vacca et al prospectively enrolled 46 CT-infected female adolescents from health centers at 2 public high schools in New York City at the time they presented for treatment and offered them EPT with patient-delivered partner therapy for their male sexual partners [80]. Subjects had urine CT NAAT at 3- and 6-month follow-up visits. Of 37 subjects completing the 3-month follow-up visit, 17 (45.9%) delivered EPT to their partner. The reinfection rate was lower in subjects who delivered EPT (3 [17.6%] vs 8 [40%]; P=.08). The reinfection rate did not differ between the groups at the 6-month visit.

Of note, there was a study by Golden et al identified in the current literature review that conducted a stepped-wedge, community-level randomized trial of an EPT intervention (patient-delivery partner therapy [PDPT] and partner services) vs control (ongoing partner services without the EPT intervention) in local health jurisdictions in Washington state [81]. Their study findings suggested that the EPT intervention may have reduced CT positivity in women by about 10% at the population level (CT positivity prevalence ratio = 0.89 [95% CI, .77–1.04]; P =.15) and led to an increase in PDPT provision by diagnosing clinical providers.

In summary, 2 studies add additional evidence that use of EPT for treatment of partners of persons with chlamydia can lower rates of repeat CT detection after treatment [79, 80]; most of the patients studied were women. Another study suggested that EPT could lower CT positivity in women at the population level [81]. Because the reinfection rate is still high with the use of EPT, other measures for preventing CT reinfection are needed.

Quality of evidence: Moderate

Strength of recommendation: B

Recommendation: EPT should continue to be recommended as a strategy for treatment of partners in persons with chlamydial infection.

Notes

Financial support. W. M. G. is supported by the Alabama–North Carolina STD/HIV Training Prevention Center (grant number 1U62PS004582) from the Centers for Disease Control and Prevention, Department of Health and Human Services, and the Division of Public Health Services. J. S. H. is supported by a National Health and Medical Research Council Research Fellowship (award number 1136117). T. D. is supported by the National Institute of Allergy and Infectious Diseases (NIAID) (grant/award numbers R01AI119164 and U19AI144181.

Supplement sponsorship. This supplement is sponsored by The Centers for Disease Control and Prevention.

Potential conflicts of interest. W. M. G. reports personal fees and research funding from Hologic, Inc, outside the submitted work. All other authors report no potential conflicts of interest.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.