Performance and acceptability of self-collected specimens for diagnosis of rectal and pharyngeal Chlamydia trachomatis and Neisseria gonorrhoeae infections among men who have sex with men in China: a randomized controlled trial Authors: Ting-Ting Jiang, Ning-Xiao Cao, Wei Luo, Zheng Li, Zheng-Qiu Huang, Tian-Jian Jia, Li Yan, Mei-Qin Shi, Yan Han, Yue-Ping Yin, Xiang-Sheng Chen Published: 2024 ABSTRACT AbstractBackgroundExtragenital Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) infections are prevalent among men who have sex with men (MSM). Self-sampling could potentially eliminate barriers to extragenital CT/NG testing for MSM that are hard to reach, who refuse to go for clinician-based testing, or who decline an examination. However, the required evidence to determine whether self-collected specimens are as accurate as clinician-taken specimens in terms of CT/NG diagnostic accuracy was limited in low and middle income countries. We therefore compared self-collected rectal and pharyngeal specimens with clinician-taken specimens for diagnostic accuracy among MSM in China.MethodsThis was a prospective convenience sample from 6 sexually transmitted infection (STI) clinics in China. We randomized the order of self-collected and clinician-taken specimens from the pharynx and rectum, plus first-void urine, for CT/NG detection. Self-sampling performance was compared with clinician-sampling as to agreement, sensitivity, and specificity. The acceptability of self-sampling was evaluated by questionnaire.ResultsAmong the 325 participants, prevalences of rectal CT and NG infections were 13.6% and 5.2% and pharyngeal CT and NG prevalences were 1.5% and 2.8%, respectively. The agreements between the CT tests with the self-collected and clinician-taken specimens were 98.8% (κ = 0.95, 95% CI 0.89-1.00) for rectal site and 99.4% (κ = 0.83, 95% CI 0.60-1.00) for pharyngeal site; and the agreements between NG tests were 99.4% (κ = 0.94, 95% CI 0.86-1.00) for rectal site and 98.2% (κ = 0.72, 95% CI 0.50–0.93) for pharyngeal site. The sensitivity and specificity of self-collected swabs was as follows: rectal CT: 93.0% and 99.6%; pharyngeal CT: 100.0% and 99.4%; rectal NG: 100.0% and 99.4%; pharyngeal NG: 88.9% and 98.4%. Self-collection was highly acceptable, showing that 62.0% MSM preferred self-sampling over clinician-sampling; 90.2% would use self-sampling for detection of CT and NG again.ConclusionsExtragenital screening for CT and NG should be recommended as part of STI services to MSM population. Self-collection of rectal and pharyngeal specimens had good performance for CT and NG tests and acceptability to the target population.Trial registrationThis trial was registered with the Chinese Clinical Trial Registry, ChiCTR2300073473. Registered 12/07/2023.Keywords: Chlamydia trachomatis, Neisseria gonorrhoeae, Self-sampling, Performance, Acceptability Background Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) are highly prevalent curable sexually transmitted infections (STIs) that can lead to serious sequelae if untreated among men and women [1]. Men who have sex with men (MSM) constitute a group at high risk for CT and NG infections, and the majority of infections in MSM are extragenital due to their engagement in oral and anal sex practices [2]. Studies have reported prevalences of pharyngeal infection among MSM ranging from 0 to 3.6% for CT and 0.5–16.5% for NG, rectal infection ranging from 2.1 to 23.0% for CT and 0.2–24.0% for NG [3]. Many guidelines have recommended extragenital testing for CT and NG among MSM [4–7]. However, these guidelines face challenges in their implementation because of poor access of this population to the clinic-based services or hesitation in collecting the specimen by clinicians. A systematic review and meta-analyses based on the studies in high-income countries (HICs) indicates a mixed sensitivity of the self-sampling, and good acceptability with high linkages to care, and recommends additional evaluations in low- and middle-income countries (LMICs) [8]. We present results of a trial to determine performance of self-collected rectal and pharyngeal specimens in comparison with clinician-taken specimens in testing for CT and NG and a survey of acceptability to self-sampling among MSM in China. Methods Study design and population This is a randomized clinical trial to evaluate the agreement of self-collected and clinician-taken rectal and pharyngal specimens in detection of CT as primary study outcome. As CT and NG detection was combined into one test, the agreement of the two sampling strategies in NG detection was evaluated as well. A convenience sample of MSM (men who engage in sexual activity with other men, regardless of how they identify themselves) aged ≥ 16 years who presented to STI clinic in Beijing (Beijing Jingcheng Skin Hospital), Nanjing (Institute of Dermatology, Chinese Academy of Medical Sciences), Wuxi (Wuxi Center for Disease Control and Prevention), Kunming (Kunming Center for Disease Control and Prevention), Fuzhou (Dermatology Hospital of Fuzhou) and Nanning (Dermatology Hospital of Guangxi Zhuang Autonomous Region) between June 2022 and June 2023 were enrolled. Patients who wished to be tested for CT and NG regardless of extragenital sex (oral or anal sex) and agreed to self-collect pharyngeal and rectal swabs in addition to standard clinician-taken swabs were invited to participate. Questionnaire surveys There were two questionnaire surveys. After the participants provided informed consent and agreed to participate in this study, they were invited to provide information of socio-demographic characteristics (age, education level, and marital status), sexual behaviors (sexual role, and number of sexual partners during the previous 6 months), history of STIs (self-reported HIV status, and prior infection with syphilis) and their perception of the advantages and disadvantages of self-sampling through a mobile phone app-based platform to complete the first questionnaire survey. The second questionnaire survey was conducted after the participants completed the specimen collections to obtain their feedback or comments regarding experience of self- and clinician-specimens collections. Randomisation and masking After completing the first questionnaire, an individual QR code was generated. Then, the study nurse scanned the individual QR code and the bar code on the sampling package. Two sets of specimens were collected. In addition to collection of the first-void urine (FVU) by the participants in the routine care, rectal and pharyngeal specimens were collected by participants themselves and clinicians separately with the recommended swabs. To minimize learning bias or any impact of collection order on nucleic acid amplification test (NAAT) performance, the order of self- and clinician-collected swabs was randomized by a statistician who had no other role in the study using computer-generated simple randomization. The order was fully protected using a sealed envelope allocated to the study number until participants had consented, provided baseline data, and enrolled. Because of the nature of the intervention, neither clinicians nor participants could be masked to the intervention, but study investigators and researchers entering and analysing data were masked to the group allocation. Intervention and specimen collection Based on the collection order presented in the envelope, either the clinician collected the rectal and pharyngeal specimens first and then the participant collected their own specimens, or vice versa, where the participant collected the specimens first and then the clinician did. Participants were given pictorial instructions to guide how to collect the specimens (i.e., use the white tip of the swab to wipe the back of participant’s throat and tonsils, insert the swab 1.5 inches into the rectum and gently rub the swab in a circle) and insertion into a tube containing 4.3mL Cobas PCR Media (stabilising reagent), with mirrors to aid pharyngeal sampling. Clinician-taken swabs were taken per routine clinical practice. The duration of each sampling procedure and whether participants sought help during self-sampling procedure were accurately recorded. All specimens were temporarily stored at 4℃ for a maximum of 2 weeks at the local laboratories. They were then transported under ‘cold Chain’ condition to the Guangzhou Kingmed Diagnostics Group Medical Laboratory Center. Testings and treatments The collected specimens were tested using the Cobas 4800 CT/NG assay according to manufacturer’s instructions (Roche, Switzerland). The laboratory team were blinded to the identification of study subjects from which study city and if the specimen was either clinician-taken or self-collected. The positive amplicons for CT or NG automatically generated from the Cobas 4800 System represent a CT or NG positivity, respectively. Study participants received test results from the self-collected urine and clinician-taken swab specimens. We provided treatment service to the participants if they are positive for CT/NG results according to the national treatment guidelines: NG infection was treated with ceftriaxone 1 g as a single dose while CT infection was treated with doxycycline 100 mg orally two times per day for at least 10 days [9]. Statistical methods and sample-size calculation Sociodemographic and behavioural data were summarised using descriptive statistics, median and interquartile range (IQR) for continuous variables and proportion and its 95% confidence intervals (CIs) for categorical variables. Testing results of self-collected urine specimens and clinician-taken pharyngeal or rectal specimens were used for calculating prevalences of genital and extragenital infections, respectively. The overall CT or NG prevalence was calculated by dividing the sum of participants who had the infection at any site by the total number of participants who were tested. To evaluate performance, sensitivity and specificity of self-collected specimens for the diagnosis of CT and NG compared with clinician-sampling (gold standard) were calculated. The percent agreement and kappa values between self-collected and clinician-taken test results were determined. Descriptive statistics from the sampling procedure and acceptability questionnaire were examined. To identify potential predictors of the outcome of interest (e.g., unwilling to perform self-sampling, preference for clinician-sampling), exploratory regression model analysis (univariate followed by multivariate) was conducted. Variables that showed significant association between variables and the outcome (P ≤ 0.1) in univariate analyses were included in the multivariate regression model to explore the association of variables with the outcome. A P ≤ 0.05 was considered statistically significant. All analyses were performed with the SPSS package version 20.0 (SPSS Inc., Chicago, IL). It was anticipated that at least 42 participants with positive CT should be included into the study in order to detect a sensitivity and specificity of 95% (rejecting with 80% power and a level of significance of 0.05). Assuming the study population had an overall CT prevalence of around 13% (unpublished data) [10], more than 323 participants should be recruited. Results Study population A total of 325 MSM were enrolled in the study: 99 from Beijing, 45 from Nanjing, 30 from Wuxi, 51 from Kunming, 50 from Fuzhou and 50 from Nanning. The mean age was 34 years (range, 17–63; median, 31); 202 (62.1%) had a college education or greater; More than two-thirds of the participants (189/283) had two or more male sex partners during the previous 6 months and 28.7% (91/317) reported ever having had a STI diagnosis. Before specimen collection, we also asked the participants their perception of the advantages and disadvantages of self-sampling. Of those responded, the most cited advantage was be more private (68.6%, 223/325) and the disadvantage was the complicated self-sampling procedure they thought (51.4%, 167/325). Prevalence of CT and NG The overall prevalences of CT and NG were 16.3% (95%CI, 12.3-20.3%; 53/325) and 8.6% (95%CI, 5.5-11.7%; 28/3325), respectively. Rectal site had higher prevalence of CT and NG (CT: 13.6%, 95%CI, 9.8-17.3%, 44/325; NG: 5.2%, 95%CI, 2.8-7.7%, 17/325) than pharyngeal site (CT: 1.5%, 95%CI, 0.2-2.9%, 5/325; NG: 2.8%, 95%CI, 1.0-4.6%, 9/325). Self-sampling performance The comparisons of self-sampling with the clinician-sampling in detection of CT and NG are shown in Table 1. Table 1Detection of CT and NG by Clinician and Self-collected pharyngeal and rectal swabs in 325 MSM in ChinaClinician-SamplingN (%)Self- SamplingN (%)CT Pharyngeal5 (1.5)7 (2.2) Rectal43 (13.2)41 (12.6)NG Pharyngeal9 (2.8)13 (4.0) Rectal17 (5.2)19 (5.8) Detection of CT and NG by Clinician and Self-collected pharyngeal and rectal swabs in 325 MSM in China Concordance between the self-collected pharyngeal specimens and the clinician-taken specimens was seen in 5 participants for CT and discordance was in 2 participants (2 positive for self-sampling but negative for clinician-sampling). For detection of NG, pharyngeal test results concordance was found in 8 participants and discordance in 6 participants (5 positive only for self-sampling, 1 positive only for clinician-sampling). Using the testing results determined by the clinician-taken pharyngeal specimens as the comparators, the sensitivity and specificity of self-collected pharyngeal specimens were 100% (95%CI, 47.8-100%) and 99.4% (95%CI, 97.8-99.9%), respectively for detection of CT and 88.9% (95%CI, 51.8-99.7%) and 98.4% (95%CI, 96.4-99.5%), respectively for detection of NG. The percentage agreement and Cohen’s kappa coefficient between the self-collected pharyngeal specimens and clinician-taken specimens were 99.4% (95%CI, 97.8-99.9%) and 0.83 (95%CI, 0.60-1.00) for detection of CT, and 98.2% (95%CI, 96.0-99.3%) and 0.72 (95%CI, 0.50–0.93) for detection of NG. When comparing the CT and NG results between the self-collected rectal specimens and clinician-taken specimens, discordance was seen in 4 participants for CT (1 positive for self-sampling but negative for clinician-sampling, 3 positive for clinician-sampling but negative for self-sampling) and 2 participants for NG (2 positive only for self-sampling). The sensitivity and specificity of self-collected rectal specimens were 93.0% (95%CI, 80.9-98.5%) and 99.6% (95%CI, 98.0-100%), respectively for detection of CT and 100% (95%CI, 80.5-100%) and 99.4% (95%CI, 97.7-99.9%), respectively for detection of NG. The percentage agreement and Cohen’s kappa coefficient between the self-collected rectal specimens and clinician-taken specimens were 98.8% (95%CI, 96.9-99.7%) and 0.95 (95%CI, 0.89-1.00) for detection of CT, and 99.4% (95%CI, 97.8-99.9%) and 0.94 (95%CI, 0.86-1.00) for detection of NG. Acceptability We received acceptability questionnaires from 266/325 (81.8%) study participants. Of all respondents, 90.2% (240/266) would be willing to self-collect the specimens for detection of CT and NG again in the future. To respond to preference of sampling options, 62.0% (165/266) of MSM would prefer self-sampling, whereas 32.7% (87/266) of MSM would prefer clinician-sampling. 86.5% of MSM (230/266) considered self-sampling procedure easy to follow and 91.4% (243/266) felt the test instructions clear. About one-third (94/266) of the participants experienced a significant or very significant discomfort in self-collection of rectal swabs but this experience was less encountered in pharyngeal swab collection (15.4%, 41/266). Preference for either sampling method did not differ by age, education level, marital status, sexual role, number of sexual partners during the previous 6 months, history of STIs. Multivariable logistic regression analysis found lower education level to be associated with unwillingness to use self-sampling again (AOR 2.90, 95%CI 1.19–7.07) (Table 2). Table 2Multivariable logistic regression analysis of the factors associated with unwillingness to use self-sampling againVariablesUnwillingness to use self-sampling againN (%)Adjusted OR (95% CI)P valueEducation level High school or below16 (61.5)2.90 (1.19–7.07)0.020 College or above10 (38.5)1Had history of STIs during the previous 6 months Yes22 (88.0)10.273 No3 (12.0)2.90 (1.19–7.07)Cited self-sampling is easier Yes7 (26.9)10.349 No19 (73.1)1.67 (0.57–4.88)Expressed significant discomfort to self-collect pharyngeal swab No19 (76.0)10.772 Yes6 (24.0)1.67 (0.57–4.88)Expressed significant discomfort to self-collect rectal swab No12 (46.2%)10.751 Yes14 (53.8%)1.19 (0.40–3.51)STIs: sexually transmitted infectionsOR: odds ratioCI: Confidence Intervals Multivariable logistic regression analysis of the factors associated with unwillingness to use self-sampling again STIs: sexually transmitted infections OR: odds ratio CI: Confidence Intervals Time cost analysis Clinicians and participants took similar time (1 min) to take the pharyngeal swabs. However, participants took longer to take their rectal swabs than clinicians (1.43 vs. 1.12). Only 22 participants (6.8%, 22/325) sought help during self-sampling procedure, including 6 for self-collection of pharyngeal swabs and 16 for rectal swabs. Discussion This is the first randomized trial to evaluate performance of self-collected rectal and pharyngeal specimens in comparison with clinician-taken specimens in testing for CT and NG in Chinese MSM. We showed that self-sampling is an appropriate alternative to the standard approach of clinician-sampling in terms of test performance and acceptability. Agreement between self-sampling and clinician-sampling is good (all > 98%), although caution must be employed when using pharyngeal self-sampling for NG, which failed reaching the WHO-recommended target product profile (sensitivity = 90.0% and specificity = 90.0%) [11]. Previous studies have showed a good performance when rectal and pharyngeal specimens were used for testing of CT and NG with cobas 4800 assay [12, 13]. Interestingly, this study indicated that the self-collected swabs had a higher number of pharyngeal CT/NG-positive and rectal NG-positive results than clinician-taken swabs, resulting in our concern in potential contamination during self-sampling. Contamination from the participant themselves has been described in previous studies and could be a source of false-positive results [14]. In our evaluation, 12.5% (10/80) of participants whose self-collected specimens were positive for CT or NG were identified to be uninfected by the specimens taken by clinicians. There may be several possibilities to result in this contamination. One is the contamination across the anatomic sites within an individual participant. In this study, we found that 30% (3/10) of participants who had inconsistent results (positive in self-collected but negative in clinician-taken specimens) were found to be positive for CT or NG at another anatomical site, which was higher than the rate of multi-site CT or NG coinfection (20.3%, 14/69). Other possibilities to cause the contamination may include inexperience in handling the collection materials, and operational environment in which the clinic surfaces were not decontaminated before and/or after each collection [15]. Although our pictorial instructions advised participants explicitly to wash their hands before swab-taking and prevent the swab tips from touching anything in the environment, future studies are needed to explore how to further minimize the risk of false-positive test results among persons who self-collect the specimens for testing. Consistent with studies in other countries [16–18], acceptability for self-sampling among Chinese MSM was very high (90.2%), but did not always translate to preference for self-sampling if given another option (e.g., clinician-sampling). However, we did not find any factors statistically associated with preference for either sampling method. Only education level was found to be related to nonacceptance of self-sampling in questionnaire surveys. Fear of taking the swab incorrectly and doubt of accuracy of test results of self-collected swabs were also identified as concerns in this study and others [16, 19]. Therefore, patients need to be clearly informed that self-sampling is easy to perform and self-collected specimens are comparable to those taken by clinicians for the detection of CT and NG, which could potentially enhance testing access for high-risk populations. Cost was another major concern in studies reporting evaluation of self-sampling [20, 21]. Results in this study indicated that it is cheaper to have the sample taken by the clinician if they need to wait for the patient to complete self-sampling. But self-sampling will be more cost-effective if the clinician’s time can be used in other patient’s care while the patient is collecting their own swabs. Mail-in home-based self-sampling test for CT or NG infection has been available in many countries including China [22], in this scenario the time needed for swab-sampling becomes irrelevant, which would further ease the burden of limited health resources, especially in LMICs. Self-sampling can also play an important role in controlling CT and NG transmission, considering that extragenital STI are not uncommon in MSM and considering, as well, the barriers to access extragenital testing. In keeping with other studies, we identified high levels of extragenital infections in MSM [3]. Among Chinese MSM in this study, 75% (21/28) of gonorrhea and 75.5% (40/53) of chlamydia would have been missed if only urogenital tests were offered. Of course, self-sampling cannot always replace clinician evaluation and proctoscopic examination, especially among patients with suspicious anal warts or anal mucosa for lesions, discharge, or abscesses. However, application of self-sampling methods is especially valuable at home or in settings that do not offer extragenital screening for CT and NG (e.g., non-government organizations), since the self-sampling can allow more people to get private and anonymous access to testing [23, 24]. Limitations of our study are that we used clinician-taken specimens as the gold reference standard which assumes 100% accuracy. However, 100% accuracy is not a clinical reality. Moreover, the discordant specimens were not retested, so we could have missed some false-positive or false-negative results. We did not record how many MSM attending the clinic were offered to participate in the study, and thus we lacked data on enrollment percentages and reasons for nonparticipation. However, it is hoped that the evidence provided in our study could help to introduce self-collected specimens as a new testing modality in STI control programs for patients that are hard to reach, who refuse to go for clinician-based testing, or who decline a clinical examination. Conclusions In conclusion, we have demonstrated good concordance between self-collected and clinician-taken pharyngeal and rectal specimens in diagnostic accuracy for CT and NG infections among Chinese MSM. Acceptability for self-sampling was very high, and self-sampling were cheaper than clinician-sampling if not overseen by the clinician or performed at home. These findings have important implications for policymakers and providers of STI services as self-sampling could be used to expand extragenital STI testing beyond healthcare facilities to non-government organizations and home-based services, which could address obstacles associated with conventional clinic-based services. Acknowledgements The authors thank MSM community organizations for helping in community mobilization and MSM recruitment; all the subjects who participated in the survey; Shanghai Zhisen Information Technology Co., Ltd. for development of the online questionnaire platform and Mr. Dong Xiao for his valuable comments and suggestions. Abbreviations Chlamydia trachomatis Neisseria gonorrhoeae Men who have sex with men Sexually transmitted infection High-income countries Low- and middle-income countries First-void urine Nucleic acid amplification test Interquartile range Confidence intervals Adjusted odds ratio World Health Organization Author contributions XSC conceived the study. XSC, TTJ, MQS, YH, and YPY designed the study. NXC, WL, ZL, ZQH, TJJ, and LY gathered the data and samples in the study sites. TTJ did primary data analysis and NXC and YH contributed to part of the analysis. TTJ wrote the first draft of the manuscript. XSC and YPY made a critical revision of the drafts. All authors read and approved the final manuscript. Funding This work was supported by the Key research and development program of Yunnan Province Science and Technology Department [202403AC100011], Chinese Academy of Medical Sciences Initiative for Innovative Medicine [2016-I2M-3-021] and the National Major Science and Technology Projects of China [2018ZX10101-001]. The funders had no role in the study design, publication decision, or manuscript preparation. Data availability Data are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate This study and survey were approved by the Medical Ethics Committee of the Chinese Academy of Medical Sciences (CAMS) Institute of Dermatology in China (approval number 2022-KY-001). Survey responses were anonymous and participation was voluntary, and informed consent was obtained from all subjects and/or their legal guardian(s). The study was conducted according to the Declaration of Helsinki. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement Data are available from the corresponding author on reasonable request.