High prevalence of rectal chlamydia and gonorrhea among men who have sex with men who do not engage in receptive anal sex

Abstract

Background:

In the United States, annual screening for rectal gonorrhea and chlamydia is only recommended for men who report receptive anal sex (RAS), but other behaviors (e.g., rimming) may lead to rectal GC/CT acquisition.

Methods:

We enrolled individuals assigned male sex at birth who reported sex with men and denied RAS in the past 2 years or reported RAS 1–2 years ago but were tested and treated since last RAS. Participants enrolled in-person at the Sexual Health Clinic in Seattle, Washington (December 2019-July 2022) or online (July 2021-March 2022). Participants completed a survey which asked about 13 non-RAS behaviors and self-collected a rectal swab for gonorrhea/chlamydia nucleic acid amplification testing (NAAT). We used log binomial regression to estimate the prevalence of rectal gonorrhea/chlamydia (adjusted prevalence ratio [aPR]) by behavior, adjusting for all other behaviors.

Results:

We enrolled 292 participants (247 in-person and 45 online); 277 (95%) had NAAT results. Rectal gonorrhea/chlamydia test positivity was 14.1% overall; 10.5% for rectal chlamydia and 4.3% for rectal gonorrhea. Most (70%) participants reported ≥1 behavior that involved direct contact with their anus. We observed a higher risk of rectal chlamydia for those who did vs. did not report peri-anal play at 12 months (aPR=2.39; 95% CI=1.10–5.22) and 2 months ((aPR=2.21; 95% CI=1.02–4.79). This was the only behavior significantly associated with testing positive.

Conclusions:

Rectal CT/GC prevalence was high among men who deny RAS, suggesting other possible routes of acquisition. Rectal screening for those who deny RAS should be made with careful consideration of individual- and population-level effects.

SHORT SUMMARY

Among MSM who do not report receptive anal sex, 70% reported at least one behavior that involved direct contact with their anus and 14.1% tested positive for rectal gonorrhea/chlamydia.

INTRODUCTION

Rectal Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (GC) infections are common among men who have sex with men (MSM), with prevalences of 16.1% and 14.8%, respectively, among sexual health clinic populations of MSM.¹ Rectal CT and GC are typically asymptomatic, but they increase the risk of HIV acquisition and transmission and likely contribute to sustained population-level transmission of CT and GC.^2–4

The United States Centers for Disease Control and Prevention (CDC) recommends screening MSM for CT/GC at exposed anatomic sites, where exposure is defined as penetration with a penis. In clinical practice this translates to only screening MSM for rectal CT/GC if they report receptive anal sex (RAS). However, in a recent US study where clinic patients were screened for rectal CT/GC regardless of reported behaviors, 32% of all rectal CT cases and 40% of all rectal GC cases among MSM were among patients who did not report RAS.⁵ Further, a number of studies have found that behaviors other than RAS (e.g., rimming, use of sex toys, fingering) may lead to rectal CT/GC acquisition.^6–8 However, most of these estimates were obtained from populations who also report RAS; thus, it is difficult to determine how common anal exposures are among men who do not engage in RAS, and which specific behaviors other than RAS may lead to rectal CT/GC acquisition.

The objectives of this study were to determine the prevalence of behaviors that may lead to rectal CT/GC acquisition and the association between these behaviors and testing positive for rectal CT/GC, among a population of MSM who do not engage in RAS.

METHODS

Study Design, Population, and Setting

“Bottoms Up” was a cross-sectional study of MSM recruited via two mechanisms: in-person and online. In-person recruitment occurred at the Public Health – Seattle & King County Sexual Health Clinic (SHC). We began in-person recruitment in December 2019, stopped in mid-March 2020 due to COVID-19 restrictions, and resumed in October 2020 through July 2022. Per routine clinical practice at the SHC, patients complete a clinical intake form which asks about their sex at birth and gender, age, gender of sex partners, and sexual behaviors in the past 12 months. Participants who reported in the intake form that they were male sex at birth, at least 16 years old, reported sex with a cisgender man in the past 12 months, and did not report RAS in the past 12 months were approached by study staff for enrollment. Staff verbally explained the study and, if patients were interested, verbally verified the date of the patient’s last RAS. If the patient had never had RAS or had RAS >2 years ago, they were eligible to enroll. If the patient reported RAS 1–2 years ago, study staff asked permission to review the patient’s SHC medical record to determine if the patient had been screened and treated for rectal CT/GC since their last RAS. Patients were only eligible to enroll if they had never had RAS, had RAS >2 years ago, or had RAS 1–2 years ago and were screened and treated since their last RAS and had not had RAS since then.

Online recruitment started in July 2021 and was active for a total of 5 months. We recruited participants online via a geospatial social networking app and the social media platforms Facebook and Instagram, as well as third-party apps and websites with which Facebook partners. We placed image-based study advertisements and text-based pop-up advertisements on these apps and sites, and we geo-targeted the advertisements to residents of King, Pierce, and Snohomish Counties (the general catchment area of the SHC) who reported being at least 18 years old. Upon clicking on the survey advertisement, individuals were taken to a survey landing page which described the purpose of the study. Individuals who chose to proceed were asked to complete a brief eligibility screener to ascertain their age, birth sex, zip code, gender of sex partners in the past year, and history of RAS in the past 2 years. Individuals who reported being male sex at birth, at least 18 years old, had sex with a man in the past 12 months, did not report RAS in the past 2 years, and whose residential zip code (self-reported) was in King, Pierce, or Snohomish counties were eligible to participate. Eligible individuals were taken to an electronic consent page.

All recruitment, consent, and study procedures were reviewed and approved by the University of Washington Institutional Review Board (IRB #00007226).

Study Procedures and Measures

All participants completed an electronic behavioral survey and self-collected a rectal specimen for CT/GC nucleic acid amplification testing (NAAT; Aptima Combo-2, Hologic Inc., San Diego, CA). The survey was programmed in REDCap^9,10 and took participants about 10–20 minutes to complete. Prior to implementing the survey, we conducted cognitive interviews with six patients from the SHC to refine the survey. The goal of the cognitive interviews was to ensure the questions would be understood by the study population and were being answered as intended. We refined the survey after each cognitive interview round in an iterative process. The final survey included general demographic questions (e.g., race and ethnicity, gender identity) and detailed sexual behavior questions. Table 1 provides the behaviors and definitions included in the survey. We asked participants if they had ever engaged in several sexual behaviors. Those who indicated ever engaging in a behavior were asked if they had engaged in those behaviors in the past 12 and 2 months, and also the date that they last engaged in those behaviors. For questions focused on the past 12 or 2 months, we only included behaviors that could lead to rectal CT/GC acquisition. These included behaviors that involved direct contact with a participant’s anus/rectum (being rimmed, being fingered, use of sex toys in their rectum, being fisted, watersports) and behaviors that did not involve direct contact with the participant’s anus/rectum (kissing, giving oral sex to someone with a penis [fellatio] or vagina (cunnilingus), rimming a partner, and felching a partner). We included these latter behaviors based on the hypothesis, specifically for CT, that C. trachomatis introduced orally can survive passage through the upper gastrointestinal tract and subsequently infect rectal tissue.^11–14

Table 1.

Behaviors included in survey and the explanation/description that appeared in the “Ever” engaged in behaviors section

Behavior	Explanation/Description in Survey
Deep Kissing	Have you EVER engaged in deep kissing? This is also called “french kissing” and is when you touch tongues with another person while kissing, and there is saliva exchanged
Insertive anal sex	Have you EVER had insertive anal sex? This is also called topping and is when you put your penis in your partner’s anus (butt)
Receptive oral sex (i.e., received oral sex)	Have you EVER performed oral sex on a man (i.e., given a man a blow job)? This is when you put your mouth/tongue on a man’s penis
Cunnilingus	Have you EVER performed oral sex on a woman? This is also known as cunnilingus. This is when you put your mouth/tongue on a woman’s vagina
Rimming someone	Have you EVER rimmed someone? Rimming is when you put your mouth and/or tongue on someone’s anus
Been rimmed	Has anyone EVER rimmed you? Being rimmed is when someone puts their mouth and/or tongue on your anus
Perianal play (Non-penetrative anal play)	Have you EVER engaged in non-penetration anal play with your anus and your partner’s penis? This is when your partner touches your anus with his penis without full penetration or when your partner’s penis enters your anus a little bit but does not completely penetrate. If you have only engaged in anal play with your penis and your partner’s anus, please mark “No” for this question.
Sex toys	Have you EVER use sex toys in your anus with a partner? This includes using butt plugs and dildos. If you have only used sex toys with a partner in their anus, please mark “No” for this question.
Fingering	Has anyone EVER fingered your anus? Being fingered in the anus is when someone puts their fingers on or in your anus. This is different than fisting, where someone puts their entire hand in your anus.
Fisting	Has anyone EVER fisted you? Being fisted is when someone puts their hand in your anus.
Watersports	Has anyone EVER urinated (peed) in your anus? This is a type of watersports.
Felching	Have you EVER felched? Felching is when you suck semen out of someone’s anus or vagina after ejaculation.

All participants self-collected a rectal swab for CT/GC NAAT. For participants recruited online, we mailed a study kit in a bubble padded envelope, which included: an Aptima^® Multitest Swab Specimen Collection Kit, a visual guide for rectal self-collection, a pre-paid mailer for participants to ship the specimen back to us, and a QR code for participants to scan to take a brief electronic survey asking them about the sexual behaviors they had engaged in since they took the main study survey. This QR code was unique for each participant and automatically linked with their original survey record. This additional survey was added for online study participants to ensure we included all behaviors the participant may have engaged in between completing the main survey and collecting the rectal specimen. We sent up to two reminder emails to participants to whom we sent a specimen collection kit but had not received it back. All rectal NAAT specimens were tested within 5 business days of receipt.

We told all participants that we would inform them of their rectal test result only if the result was positive. For those testing positive for rectal CT, we offered to call the prescription (100 mg doxycycline twice daily for 7 days) into the participant’s preferred pharmacy or to have them come to the SHC to pick up the prescription. We asked participants who tested positive for rectal GC to come to the SHC for treatment (ceftriaxone 500 mg intramuscular in a single dose) and for collection of a rectal specimen for GC culture, per our clinic’s standard of care.

Participants recruited in-person were paid $25 cash for participating in the study. Participants recruited online were paid a $25 Amazon.com gift card after receipt of their specimen collection kit.

Definitions and Statistical Analyses

For the entire study population, we report the prevalence of ever engaging in rectal behaviors and engaging in behaviors in the past 12 and 2 months. For those with a NAAT result, we report the behavior-specific rectal CT and GC test positivity, comparing test positivity among those who did vs. did not report the behavior. We used log binomial regression with robust standard errors to estimate the prevalence ratio (PR) of rectal CT or GC comparing those who did vs. did not report a behavior, adjusting for all other behaviors as well as adjusting for the a priori confounders age (continuous) and number of sex partners in the past 12 months (continuous). The adjusted PR’s provide the independent association between a specific behavior and the prevalence of rectal CT or GC. For analyses with an outcome of CT/GC test positivity, we created a single behavior variable called “non direct contact anal behaviors” that included the following behaviors that did not involve direct contact with the participant’s anus/rectum: giving oral sex to someone with a penis (fellatio) or vagina (cunnilingus), rimming a partner, and felching a partner. All analyses were completed in Stata Version 16.1 (StataCorp LLC, College Station, TX) and at an alpha level of 0.05.

RESULTS

During the 21-month in-person recruiting period, we enrolled 258 participants in-person. Eleven of these individuals reported in the research survey that they had RAS in the past 12 months and were excluded from further analyses. Of these 247 in-person participants, 244 had NAAT results available. During the 5-month online recruiting period, 1,077 individuals clicked on our study advertisement and started the eligibility screener; 766 (71%) answered all eligibility questions and 633 (83%) were ineligible to participate: 4 (<1%) were <18 years old, 15 (2%) lived outside the study area, 15 (2%) were not assigned male at birth, 109 (17%) did not have sex with a man in the past year, and 534 (85%) had receptive anal sex in the past 2 years; these categories are not mutually exclusive. Of the 133 (17%) who were eligible, 99 (74%) consented and started the study. Just over half (n=54) completed the survey and 52 provided a valid address for us to send a test kit; however, 7 of these individuals reported in the research survey that they had RAS in the past 12 months and were excluded from further analyses. Of the 45 individuals who were sent a test kit and were eligible, 33 (75%) returned their test kit; all had NAAT results available. Thus, our final analytic sample sizes combining those recruited in-person and online were 292 for behavioral analyses and 277 for analyses that included NAAT results.

Just over half of the 292 participants were under 35 years old, 55% self-reported being white race, 21% self-reported Hispanic or Latinx ethnicity, and the majority identified as male gender (Table 2). The median number of sex partners in the past 12 months was four; 44% reported five or more sex partners in the past 12 months. The characteristics of the 277 individuals with NAAT results were nearly identical to the distributions presented in Table 2 (data not shown).

Table 2.

Characteristics of Study Population Enrolled in Bottom’s Up (N = 292)

Characteristic	N (%)
Age
16–24	33 (11)
25–29	55 (19)
30–34	69 (24)
35–44	69 (24)
45+	66 (23)
Not recorded	0 (0)
Race
American Indian or Alaska Native	12 (4)
Asian	40 (14)
Black or African American	50 (17)
Native Hawaiian or Pacific Islander	8 (3)
White	162 (55)
Not recorded	20 (7)
Hispanic or Latinx Ethnicity
Yes	60 (21)
No	226 (77)
Not recorded	6 (2)
Gender Identity
Male	277 (95)
Female	1 (<1)
Non-binary / genderqueer	9 (3)
Not recorded	5 (2)
Education
Less than High School	4 (1)
High school or equivalent	92 (32)
Associate’s Degree	31 (11)
Bachelor’s Degree	114 (39)
Graduate or professional degree	48 (16)
Not recorded	3 (1)
Number of sex partners, past 12 months
1	50 (17)
2–4	105 (36)
5–9	51 (17)
10+	74 (25)
Not recorded	12 (4)

Reported behaviors are described in Table 3. The four most common behaviors that participants reported ever engaging in were: deep kissing (98%), giving oral sex to a man (98%), insertive anal sex (96%), and rimming someone (86%). The most commonly reported behaviors at 12 and 2 months were giving oral sex to a man (80% and 67%, respectively) and rimming someone (66% and 54%, respectively).

Table 3.

Number and percentage of respondents who reported engaging in behaviors ever and in the past 12 and 2 months (N = 292)

Behavior	Ever N (%)	Past 12 months N (%)	Past 2 months N (%)
Deep Kissing	279 (98)	250 (87)	220 (77)
Receptive Anal Sex	200 (70)	0 (0)	0 (0)
Insertive Anal Sex	276 (96)	-- *	-- *
Gave oral sex to man (fellatio)	281 (98)	228 (80)	190 (67)
Given oral sex to woman (cunnilingus)	112 (39)	33 (11)	22 (8)
Rimmed Someone	247 (86)	189 (66)	154 (54)
Been Rimmed	257 (90)	146 (51)	93 (33)
Peri-Anal Play	203 (70)	119 (41)	78 (27)
Sex Toys	109 (38)	28 (10)	15 (5)
Been fingered	216 (75)	79 (28)	46 (16)
Been fisted	3 (1)	1 (<1)	0 (0)
Watersports	9 (3)	1 (<1)	0 (0)
Felching	17 (6)	10 (3)	6 (2)

^*Survey did not ask about insertive anal sex in the past 12 and 2 months because it is not a behavior that would lead to rectal GC/CT acquisition for the respondent

Seventy percent of participants (n=203) reported at least one behavior that involved direct contact with the participant’s anus/rectum in the past 12 months (i.e., being rimmed, peri-anal play, being fingered, sex toys), and 50% (n=146) reported at least one direct anal/rectal contact behavior in the past 2 months. At 12 months, 87% (n=254) of participants reported at least one behavior that did not involve direct contact with their anus/rectum (i.e., giving oral sex, rimming someone, felching someone). At 2 months the analogous percentage was 78% (n=227).

The overall rectal CT/GC test positivity was 14.1% (39 of 277), with 10.5% (29 of 277) of participants testing positive for rectal CT and 4.3% (12 of 277) testing positive for rectal GC. Overall, rectal CT test positivity did not substantially differ by whether or not participants reported specific behaviors, except for peri-anal play (Table 4). Adjusting for all other behaviors, participants reporting peri-anal play in the past 12 months had a 2.4-fold higher prevalence of rectal CT compared to those who did not report peri-anal play (aPR=2.39; 95% CI=1.10 – 5.22). We did not observe significant differences in rectal GC test positivity for any behaviors reported in the past 12 months (Table 4).

Table 4.

Association between reported behaviors at 12 months and testing positive for rectal GC or CT (N=277)

	Rectal CT		Rectal GC
Behavior	N (%) Testing Positive for Rectal CT*	Adjusted** Prevalence Ratio (95% CI)	N (%) Testing Positive for Rectal GC*	Adjusted** Prevalence Ratio (95% CI)
Non-contact behaviors†
Yes (N=242)	3 (9.1)	1.14 (0.28 – 4.63)	10 (4.1)	0.42 (0.10 – 1.76)
No (N=33)	25 (10.3)	1.0	2 (6.1)	1.0
Been Rimmed
Yes (N=138)	12 (8.7)	0.60 (0.29 – 1.25)	6 (4.4)	0.93 (0.29 – 2.95)
No (N=133)	16 (12.0)	1.0	6 (4.5)	1.0
Peri-Anal Play
Yes (N=112)	16 (14.3)	2.39 (1.10 – 5.22)	5 (4.5)	0.97 (0.27 – 3.42)
No (N=161)	12 (7.5)	1.0	7 (4.4)	1.0
Sex Toys
Yes (N=27)	2 (7.4)	0.95 (0.24 – 3.71)	1 (3.7)	1.21 (0.19 – 7.80)
No (N=244)	26 (10.7)	1.0	11 (4.5)	1.0
Been fingered
Yes (N=74)	6 (8.1)	0.63 (0.28 – 1.43)	2 (2.7)	0.56 (0.13 – 2.45)
No (N=199)	22 (11.1)	1.0	10 (5.0)	1.0

CT, Chlamydia trachomatis; GC, Neisseria gonorrhoeae; CI, confidence interval

^*N is number who tested positive for rectal CT or GC; % is percent who tested positive for rectal CT or GC out of those who did or did not report the behavior (row-specific N’s provided in far-left column)

^**PR adjusted for all other behaviors in table, age, and number of sex partners in the past 12 months

^†Includes giving oral sex to man (fellatio), giving oral sex to a woman (cunnilingus), rimming someone or felching. Categorized as “yes” is respondent reported any of these behaviors and “no” if they reported none of them

Table 5 describes test positivity by behaviors in the past 2 months. Similar to behaviors in the past 12 months, we found that the only behavior significantly associated with rectal CT test positivity was peri-anal play (aPR=2.21; 95% CI=1.02 – 4.79). For rectal GC, though we did not observe any statistically significant differences in test positivity by behavior reported in the past 2 months, we noted that those who reported being rimmed, peri-anal play, and use of sex toys had a notably higher prevalence of rectal GC compared to those who did not report those behaviors.

Table 5.

Association between reported behaviors at 2 months and testing positive for rectal GC or CT (N=277)

	Rectal CT		Rectal GC
Behavior	N (%) Testing Positive for Rectal CT*	Adjusted** Prevalence Ratio (95% CI)	N (%) Testing Positive for Rectal GC*	Adjusted** Prevalence Ratio (95% CI)
Non-contact behaviors†
Yes (N=215)	19 (8.8)	0.46 (0.20 – 1.10)	10 (4.7)	0.81 (0.18 – 3.62)
No (N=60)	9 (15.0)	1.0	2 (3.3)	1.0
Been Rimmed
Yes (N=87)	7 (8.1)	0.57 (0.24 – 1.32)	5 (5.8)	1.09 (0.33 – 3.61)
No (N=184)	21 (11.4)	1.0	7 (3.8)	1.0
Peri-Anal Play
Yes (N=72)	11 (15.3)	2.21 (1.02 – 4.79)	4 (5.6)	1.26 (0.38 – 4.20)
No (N=201)	17 (8.5)	1.0	8 (4.0)	1.0
Sex Toys
Yes (N=14)	2 (14.3)	1.54 (0.35 – 6.86)	1 (7.1)	2.01 (0.32 – 12.7)
No (N=257)	26 (10.2)	1.0	11 (4.3)	1.0
Been fingered
Yes (N=43)	4 (9.3)	0.90 (0.34 – 2.39)	2 (4.7)	0.84 (0.20 – 3.48)
No (N=230)	24 (10.4)	1.0	10 (4.4)	1.0

CT, Chlamydia trachomatis; GC, Neisseria gonorrhoeae; CI, confidence interval

^*N is number who tested positive for rectal CT or GC; % is percent who tested positive for rectal CT or GC out of those who did or did not report the behavior (row-specific N’s provided in far-left column)

^**PR adjusted for all other behaviors in table, age, and number of sex partners in the past 12 months

^†Includes giving oral sex to man (fellatio), giving oral sex to a woman (cunnilingus), rimming someone or felching. Categorized as “yes” is respondent reported any of these behaviors and “no” if they reported none of them

DISCUSSION

In this cross-sectional study that enrolled MSM who do not engage in RAS, 14.1% tested positive for rectal CT or GC. We found that the majority of participants reported engaging in behaviors that involve direct contact with their anus/rectum, and that some behaviors – particularly peri-anal play – may significantly increase the risk of testing positive for rectal CT. Although our findings indicate that limiting rectal screening to individuals who report RAS may miss a substantial number of rectal CT/GC infections, the clinical implications of our findings are somewhat unclear. Careful consideration of the individual- and population-level benefits of rectal screening in this population should be made prior to substantially revising national screening guidelines.

Our key finding in this study is the very high rectal CT/GC test positivity among individuals who do not report RAS. We found that 10.4% of participants tested positive for rectal CT and 4.3% tested positive for rectal GC. Because routine screening for rectal STIs in the US is limited to those who report RAS, there is a paucity of data on rectal CT/GC test positivity among those who do not report RAS, making it difficult to contextualize our findings in those of other studies or populations. However, this test positivity is close to the range of test positivity (rectal CT: 10.2% – 20.8%; rectal GC: 7.6% – 18.1%) reported from SHC’s participating in CDC’s STD Surveillance Network (SSuN) where rectal screening is generally limited to individuals who reported RAS.¹

We also found that men who do not engage in RAS report a variety of behaviors that have the potential to lead to rectal CT/GC acquisition. Seventy percent of the study population reported at least one behavior in the past 12 months that involved direct contact with the rectum and 50% reported this at 2 months. Other studies have similarly identified a high prevalence of non-RAS behaviors that involve the anus among MSM^{6–8,15–18}, but most of these studies have been among men who also report RAS. Our study demonstrates that a very high proportion of individuals who would typically be considered at low risk for rectal infections (because they do not report RAS) may be frequently exposed to CT/GC at the anus/rectum through other behaviors.

However, we did not observe clear associations between most specific behaviors and testing positive for rectal CT/GC. The exception was peri-anal play, which we found may significantly increase the risk of rectal CT above and beyond the effect of other behaviors. This is somewhat intuitive, given it is the only behavior in the survey that may involve some direct contact of semen (pre-ejaculate) on the anus or slightly inside the rectum. For rectal GC, we found that several behaviors reported at 2 months (i.e., being rimmed, peri-anal play, use of sex toys) did result in a higher rectal GC test positivity but these findings did not reach statistical significance, which may be because we were somewhat underpowered to detect associations. There is very limited information on behaviors other than RAS that may lead to rectal CT/GC acquisition. Rimming^6,7, use of sex toys^7,8, fingering^6,18, fisting^6,7, and felching⁷ have all been associated with rectal CT or GC. But most of these studies also included MSM who report anal sex, which confounds associations between individual non-RAS behaviors and CT/GC infections. Only a prospective study of Australian MSM in the early 2000’s (the Health in Men [HIM] Study) stratified their analysis by whether or not participants had also reported RAS. In that study, Jin and colleagues noted associations between rectal CT/GC and fingering, fisting, rimming, and use of sex toys among MSM who did not engage in RAS.⁶ Our findings are generally in line with that of Jin and colleagues, suggesting that these non-RAS behaviors may indeed lead to rectal CT/GC acquisition.

Although we observed a high prevalence of rectal CT/GC in this population and believe that our results support the idea that non-RAS behaviors can lead to rectal CT/GC acquisition, the implications of our findings on rectal screening guidelines are unclear. Given recent calls for decreased screening among MSM because of its potential role in driving antimicrobial resistence^19–21, we believe that any changes to increase rectal screening should be made with careful consideration of its individual and population-level impacts. At an individual-level, the majority of rectal infections are asymptomatic.²² Although they can lead to proctitis², this appears to occur in only the minority of cases. One of the main individual-level health implications of rectal CT/GC among MSM is its association with HIV acquisition and transmission.^3,4 But among MSM who do not engage in RAS, like our study population, it is unclear whether or how a prevalent rectal infection would actually increase HIV risk. At a population-level, screening MSM who do not engage in RAS for rectal CT/GC could have an impact on population-level CT/GC transmission if a large portion of this population later engages in RAS prior to the infection being treated or spontaneously resolving, or if the transmission probabilities of rectal CT/GC from the rectum to a partner’s pharynx were high (keeping in mind that in a population who does not currently engage in RAS, there is no transmission pathway from the rectum to the partner’s urethra). At present, these rectal-pharyngeal transmission probabilities are unknown but represent a major gap in our understanding of these infections, and could help inform whether or not increased rectal screening would outweigh potentially non-essential antibiotic use. Although we believe the evidence is not strong enough at this time to recommend changes to national rectal screening guidelines, we acknowledge and understand that patients and providers may have concerns about the presence of untreated CT/GC in the rectum even among men who do not engage in RAS. To that end, providers may wish to use shared-decision making in considering whether rectal screening is appropriate for some MSM who do not report RAS.

This study has a number of strengths. We recruited a study population who did not engage in RAS to eliminate the possibility that rectal infections were acquired via that route. Our study survey was developed after completing cognitive interviews with the target population, and we are confident that our survey comprehensively captured anal/rectal exposures. There are also several key limitations. First, we may have been underpowered to detect differences in test positivity by behavior. Although our analytic sample size included nearly 300 people, the total number reporting any given behavior at 12 or 2 months was relatively small. Similarly, participants engaged in a wide variety of behaviors, and we may have been underpowered to detect the independent association between any given behavior and rectal CT/GC. Second, the behavioral data are subject to misclassification due to recall bias. Participants may not have recalled if they engaged in specific behaviors or the dates in which they engaged in these behaviors. Third, although NAATs are the diagnostic gold standard for rectal CT/GC, they are not able to distinguish between “true” active infections versus detection of nonviable bacteria or cell-free nucleic acid. Thus, it is possible that some NAAT-positive specimens were not “true” infections. Fourth, we did not collect information on condom use or dental dam use for these behaviors, nor did we stratify behaviors by whether or not participants reported use of saliva for behaviors (e.g., fingering, use of sex toys) for which saliva may be the likely route of CT/GC transmission.¹⁸ The survey did include questions about saliva use for these behaviors, in addition to other details about these behaviors (e.g., number of acts). Future analyses will focus on these additional details which may help elucidate the role of these behaviors in CT/GC transmission. Fifth, our survey may not have captured all possible non-RAS behaviors. However, we developed the survey after formative work with SHC patients and we believe the survey was comprehensive and clear. Indeed, only one participant checked that they had engaged in a behavior not provided on the list (that participant indicated that they had used sex toys, and was re-categorized as such). Finally, the participants we enrolled were individuals in the Seattle metropolitan area who either attended our Sexual Health Clinic or clicked on an online advertisement, and our findings may not be generalizable to other populations.

In summary, we identified a high prevalence of rectal CT and GC among a population of MSM who do not engage in RAS. Per national screening guidelines, these individuals would not be screened for rectal infections. However, the exact clinical implications of our findings are not entirely clear, and as such we believe that any changes to increase rectal screening should be made with careful consideration of its individual and population-level impacts.