Article Text

Download PDFPDF

Original article
Sampling technique and detection rates of oropharyngeal and anorectal gonorrhoea using nucleic acid amplification tests in men who have sex with men
  1. Tim Z T Yang1,
  2. Marcus Y Chen1,2,
  3. Tim R H Read1,2,
  4. Robert Needleman1,
  5. Catriona S Bradshaw1,2,
  6. Ria Fortune1,
  7. Christopher K Fairley1,2,
  8. Eric P F Chow1,2
  1. 1 Melbourne Sexual Health Centre, Alfred Health, Melbourne, Victoria, Australia
  2. 2 Central Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Victoria, Australia
  1. Correspondence to Dr Eric P F Chow, Melbourne Sexual Health Centre, 580 Swanston Street, Carlton, VIC 3053, Australia; echow{at}mshc.org.au

Abstract

Objectives The objective of this study was to examine the associations between clinicians’ self-reported sampling technique and the detection rate of gonorrhoea at the oropharynx and anorectum using a highly sensitive nucleic acid amplification test (NAAT).

Methods We analysed oropharyngeal and anorectal gonorrhoea swab results among men who have sex with men attending the Melbourne Sexual Health Centre (MSHC) between March 2015 and December 2016. Swabs were tested by NAAT using the Aptima Combo 2 transcription-mediated amplification assay due to its high sensitivity. Clinicians at MSHC were invited to complete a questionnaire on sampling techniques in November 2016. Univariable generalised estimating equations (GEE) logistic regressions were performed to determine the association between gonorrhoea detection rates and clinicians’ sampling technique. Patients’ epidemiological risk factors were included in the multivariable GEE logistic model.

Results A total of 2605 oropharyngeal gonorrhoea and 2392 anorectal gonorrhoea swab results were analysed. There was no significant difference in the detection rates of gonorrhoea between the 23 clinicians at the oropharynx (range 3.6%–16.9%, median 8.2%, P=0.302) or and anorectum (range 2.4%–17.3%, median 10.5%, P=0.177). Variations in clinicians’ self-reported sampling technique were not associated with oropharyngeal or anorectal gonorrhoea detection rates after adjusting for patients’ epidemiological risk factors.

Conclusions This study shows that differences in clinicians’ self-reported sampling technique did not result in measurable differences in the detection rate for oropharyngeal or anorectal gonorrhoea when using NAAT.

  • gonorrhoea
  • neisseria gonorrhoea
  • gay men
  • screening

Statistics from Altmetric.com

Introduction

The incidence of Neisseria gonorrhoeae among men who have sex with men (MSM) is rising in many countries including Australia.1 2 To address this, several countries, including Australia, have recommended frequent asymptomatic screening for gonorrhoea in MSM.3 Gonorrhoea can be detected by culture or nucleic acid amplification test (NAAT). NAAT has a higher sensitivity but lower specificity compared with culture,4 5 and most countries recommend using NAAT for gonorrhoea screening in the MSM population.3 6

The precise sampling technique for taking an oropharyngeal or anorectal swab is not well defined in the literature. A previous study has shown that detection of oropharyngeal gonorrhoea by culture was greater when a gag reflex was more commonly elicited.7 Since March 2015, our centre has replaced culture with NAAT for screening due to its higher sensitivity.8–11 Given the increased sensitivity of NAAT, it was unclear whether clinicians’ sampling technique still affected the detection of gonorrhoea at extragenital sites.

The aim of this study was to determine whether detection rates of oropharyngeal and anorectal gonorrhoea in MSM were influenced by variations in clinicians’ self-reported sampling technique.

Methods

Study population

This study analysed digital records of patients attending the Melbourne Sexual Health Centre (MSHC), the main public sexual health service in Victoria, Australia. The centre operates a walk-in, triage clinic where patients are allocated to either doctors or nurses based on their symptoms and STI risk profile. We included male patients who were aged 16 years or above, self-reported having sex with another man and were triaged to see doctors between March 2015 and December 2016.

MSM attending MSHC were offered free testing for STIs including oropharyngeal and anorectal gonorrhoea. Swabs were tested for gonorrhoea by NAAT using the Aptima Combo 2 transcription-mediated amplification (TMA) assay (Hologic, San Diego, California, USA). The Aptima Combo 2 TMA assay is an especially sensitive NAAT test.12We excluded tests of cure where the same test was repeated within 30 days. Invalid and indeterminate NAAT results were also excluded. Demographic characteristics (eg, age) and epidemiological data (eg, number of male partners and condom use in the last 3 and 12 months, HIV status and self-reported contact of infection) were collected as part of routine clinical practice.

Clinicians’ sampling technique

In November 2016, 29 clinicians at MSHC were invited to complete a questionnaire about their oropharyngeal and anorectal sampling technique. Clinicians reported how frequently they swabbed various oropharyngeal anatomical sites (tonsils, oropharynx and so on) and the frequency of eliciting a gag reflex when obtaining an oropharyngeal specimen. Information about anorectal sampling technique included the distance that the swab was inserted within the anal canal and the frequency with which clinicians used certain techniques such as moistening or rotating the swab.

Statistical analyses

Oropharyngeal and anorectal gonorrhoea detection rate was calculated for each clinician by dividing the number of swabs with positive results by the total number of swabs collected by the clinician. The χ2 test was used to test whether there was any difference in detection rates of oropharyngeal and anorectal gonorrhoea between clinicians during the study period. We reported the median and IQR for continuous variables such as patients’ age, and the frequency of the sampling technique by the clinicians.

A generalised estimating equation (GEE) approach was used with an exchangeable correlation structure and a logit link function for univariable and multivariable logistic regression models to examine the association between clinicians’ sampling technique and gonorrhoea detection rate at each anatomical site (ie, oropharynx and anorectum). The continuous variables (ie, age and number of partners) were categorised into binary variables at the median to fit into the logistic regression. All clinicians’ sampling techniques were included as a continuous variables into the multivariable logistic regression models. Patients’ factors were considered as potential confounders, and factors with P<0.10 in the univariable analyses were accounted for in the multivariable model. Patient factors include age, risk profiles (ie, condom use in the last 3 or 12 months, number of male partners in the last 3 or 12 months, self-report of lifetime drug use and HIV status), contact of gonorrhoea, diagnosis of gonorrhoea at other anatomical sites (ie, anorectal or oropharyngeal) and urethral chlamydia. For oropharyngeal samples, we considered the number of partners and condom use in the past 3 months, whereas for anorectal samples, we considered the past 12 months, based on the natural duration of untreated infection.13 First-void urine was used for testing urethral chlamydia, which was used as an indicator for level of risk. Urethral gonorrhoea was not considered, because it was only tested in symptomatic patients as per the clinic guidelines. Sensitivity analyses of multivariable logistic regression models were performed to examine the effect of each sampling technique factor adjusting for patients’ characteristics only. Unadjusted and adjusted intracluster correlation coefficients (ICC) and its 95% confidence interval (CI) were calculated using the random-effects model. ICC was used to assess the strength of association between clinicians’ sampling technique and detection rate. All statistical analyses were conducted using Stata (V.14).

Results

Study population

There were 38 clinicians who provided clinical services at MSHC between March 2015 and December 2016. Twenty-nine of these clinicians were still working at MSHC in November 2016 when the questionnaire was disseminated. Of these, 25 (86.2%) completed and returned the questionnaire and 4 declined to complete the questionnaire. Two of these 25 clinicians were excluded from the analysis because they only obtained one swab each during the study period. The remaining 23 clinicians were included in the final analysis, which contributed a total of 3126 consultations that were MSM. After excluding tests of cure and invalid and indeterminate results, there were 2605 oropharyngeal gonorrhoea and 2392 anorectal gonorrhoea test results included in our analysis.

Of the 23 clinicians included in the final analysis, most clinicians swabbed both tonsils together with the oropharynx for oropharyngeal gonorrhoea detection (median=70%), and the median frequency of eliciting a gag reflex from the patients was 70% (table 1). When taking anorectal swabs for anorectal gonorrhoea, most clinicians rotated the swab within the anal canal (median=90%) and moistened the swab before insertion (median=80%).

Table 1

Summary of oropharyngeal and anorectal sampling technique among 23 clinicians at Melbourne Sexual Health Centre

Oropharyngeal gonorrhoea

The number of oropharyngeal gonorrhoea swabs obtained by each clinician ranged from 28 to 275 (median=107; IQR=65–136). Detection rates of oropharyngeal gonorrhoea ranged from 3.6% to 16.9% (median=8.2%) between clinicians, but the difference was not statistically significant (P=0.302). The multivariable analysis showed that clinicians’ swabbing techniques such as frequency of reported gag reflex and the anatomical location swabbed were not associated with oropharyngeal gonorrhoea detection by NAAT, after adjusting for patients’ epidemiological risk factors alone (online supplementary table S1) and after adjusting for patients’ characteristics and all other sampling technique factors (table 2). The ICC estimates for all sampling techniques were minuscule, and there was no significant correlation between clinicians’ sampling technique with the detection rate of oropharyngeal gonorrhoea (table 3).

Supplementary file 1

Table 2

Univariable and multivariable generalised estimating equation analyses of factors associated with the detection rate of oropharyngeal gonorrhoea

Table 3

Unadjusted and adjusted ICC for clinicians’ sampling technique with the detection rate of oropharyngeal and anorectal gonorrhoea

Anorectal gonorrhoea

The number of anorectal swabs for gonorrhoea obtained by each clinician ranged from 29 to 259 (median=97, IQR=58–129). Detection rates for anorectal gonorrhoea did not vary significantly between clinicians (P=0.177), ranging from 2.4% to 17.3%, with a median of 10.5%. The factors associated with the detection rate of anorectal gonorrhoea are shown in table 4. Clinicians’ sampling techniques such as the distance of swab inserted from the anal verge, asking patients to relax the anal sphincter, and rotation of swabs were not associated with anorectal gonorrhoea detection rate after adjusting for patients’ epidemiological risk factors. The univariable analysis suggested that high frequency of moistening swab before insertion was more likely to have a positive result for anorectal gonorrhoea; however, this association was not significant after adjusting for patients’ characteristics (online supplementary table S2) and after adjusting for patients’ characteristics and all other sampling technique factors (table 4). The ICC estimates for all sampling techniques were minuscule, and there was no significant correlation between clinicians’ sampling technique with the detection rate of anorectal gonorrhoea (table 3).

Table 4

Univariable and multivariable generalised estimating equations analyses of factors associated with the detection rate of anorectal gonorrhoea

Discussion

To our knowledge, this is the first study examining the association between clinicians’ sampling technique and gonorrhoea detection rates at extragenital sites using NAAT with adjustment for patients’ epidemiological risk factors. Our results show that differences in clinicians’ self-reported sampling technique did not result in measurable differences in the detection rate for anorectal or oropharyngeal gonorrhoea when using NAAT. This contrasts with findings of previous studies showing that sampling technique does influence oropharyngeal gonorrhoea detection when using culture,7 14 which is less sensitive than NAAT, especially using the Aptima Combo 2 TMA assay.

This study has several limitations that should be considered when interpreting our results. First, clinicians were asked to report their technique over the previous two years and hence recall bias may have occurred. Second, clinicians could only provide one overall estimate to quantify how frequently they used each technique rather than providing data on the technique used for every individual swab analysed in this study. No doubt, there would have been variations in sampling technique between patients, and we have assumed each clinician used a uniform technique to collect all of their specimens. It is possible there were differences in sampling technique between clinicians that we were unable to measure with precision. Third, patients were not randomly allocated to clinicians due to the shift patterns of the clinicians included in the analysis and thus detection rate could potentially be influenced by variability in patients. Fourth, there is no validated ‘gold standard’ test result for patient, and thus we could only examine the detection rate but not the sensitivity and specificity between clinicians.

Razali et al 7 showed that detection rates of oropharyngeal gonorrhoea by culture varied across clinicians and that a higher self-reported frequency of eliciting a gag reflex was associated with a higher detection rate of oropharyngeal gonorrhoea at our centre.7 However, we did not find a significant association between self-reported frequency of gag reflex and detection rate of oropharyngeal gonorrhoea by NAAT in this study. There are two possible reasons. First, NAAT was used in this study, and culture was used in Razali’s study. NAAT has a higher sensitivity than culture, and hence sampling technique may be less important using this detection method.8–11 Second, the median reported frequency of eliciting a gag reflex at our clinic has increased over time, from 57% in 2006–2008 reported in Razali’s study7 to 70% in 2015–2016 reported in this study. This shows that sampling technique has changed among staff at MSHC, probably as a consequence of Razali’s findings.

Another study at our centre suggested that swabbing both the tonsils and oropharynx resulted in higher bacterial loads that increased the detection rate of gonorrhoea using culture.14 Again, the greater sensitivity of NAAT could have made sampling technique less important.

The flexibility in technique allowed by NAAT testing also supports mounting evidence that self-collected oropharyngeal sampling is an acceptable alternative to clinician-collected swabs.15–18 Self-collected anorectal sampling is already a widely accepted practice.15 18 There have been very limited studies comparing clinician-collected versus self-collected samples for the detection of anorectal and oropharyngeal gonorrhoea in men. Lunny et al 19 have recently published a meta-analysis and shown that self-collected oropharyngeal swabs have a high sensitivity (0.91; 95% CI 0.75 to 0.98) and specificity (0.97; 95% CI 0.95 to 0.98) compared with clinician-collected swabs for oropharyngeal gonorrhoea.19 However, this result is based only on one available study.20 Less invasive tests such as self-collected gargle or mouth pad specimens may increase acceptability of STI testing even further.4 All in all, there does not appear to be an association between clinicians’ sampling technique and detection rate of gonorrhoea at extragenital sites using NAAT.

Key messages

  • This is the first study to investigate clinicians’ sampling technique and its effect on detection rate of gonorrhoea at extragenital sites using nucleic acid amplification tests (NAAT).

  • Sampling technique has been shown to affect detection of oropharyngeal gonorrhoea using culture, but this finding had not been studied using NAAT.

  • Sampling technique does not appear to affect detection rate of anorectal or oropharyngeal gonorrhoea when using NAAT.

Acknowledgments

We acknowledge Afrizal Afrizal for his assistance with data extraction and Sandra Walker and Alison Clough for their assistance with disseminating the questionnaires to the staff at the Melbourne Sexual Health Centre.

References

Footnotes

  • CKF and EPFC are joint last authors.

  • Handling editor Jackie A Cassell

  • Contributors EPFC, CKF, TRHR and MYC conceived the study and designed the questionnaire for the clinicians. TZTY performed the data analysis and wrote the first draft of the manuscript. All authors were involved in data interpretation, revised the manuscript critically for important intellectual content and approved the final version.

  • Funding This work was supported by the National Health and Medical Research Council (NHMRC) programme grant (number 568971). EPFC and TRHR are supported by the NHMRC Early Career Fellowships (EPFC: 1091226; TRHR: 1091536).

  • Competing interests None declared.

  • Ethics approval Ethical approval was obtained from the Alfred Hospital Ethics Committee, Melbourne, Australia (number 511/16).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.