Article Text
PDF
Abstract
Objectives Reference laboratories are increasingly using more sensitive rapid molecular techniques, such as nucleic acid amplification tests (NAATs), to diagnose infections with Neisseria gonorrhoeae. We determined the proportion of patients at sentinel genitourinary medicine clinics in England whose NAAT-positive diagnoses were also culture-positive for N. gonorrhoeae, and investigated whether they differed from those that were not.
Methods Behavioural and clinical data from all NAAT-positive patients reported from 23 clinics included in the Gonoccocal Resistance to Antimicrobials Surveillance Programme from July to September 2012 were included in this analysis. Unadjusted and adjusted associations between patient characteristics and culture-positive infection with N. gonorrhoeae were determined.
Results Of 3076 NAAT-positive patients, 46.4% had culture-positive infections. Most NAAT-positive patients were <35 years old (73.0%), white (67.9%), and men who had sex with men (60.1%). Women and men who had sex with men were less likely than heterosexual men to have culture-positive infections (adjusted OR (95% CI) 0.53 (0.41 to 0.68), p<0.001; and 0.74 (0.59 to 0.93), p=0.010, respectively), while those who were symptomatic (4.61 (3.92 to 5.42), p<0.001), and those presenting with infection at multiple sites (2.15 (1.76 to 2.62), p<0.001) were more likely to have culture-positive infections.
Conclusions Although gonococcal isolates were available from almost half of the NAAT-positive patients, culture was not attempted or may have failed in the remainder. Patients with culture-positive isolates were not representative of all NAAT-positive patients. Routine culture is necessary for monitoring emerging antimicrobial resistance and to inform gonorrhoea treatment guidelines.
- ANTIBIOTIC RESISTANCE
- GONORRHOEA
- SURVEILLANCE
Statistics from Altmetric.com
Introduction
Gonorrhoea is a sexually transmitted infection (STI) caused by the gram-negative bacterium Neisseria gonorrhoeae. It is the second most commonly diagnosed bacterial STI in England, with 29 291 diagnoses reported in 2013 from genitourinary medicine (GUM) clinics.1 Since 2009, there has been an increasing trend in diagnoses among men, and it is currently the most commonly diagnosed STI among men who have sex with men (MSM).1 Although the rise is partly due to the use of more sensitive diagnostic tests and more testing, increased transmission among MSM is likely to be contributing.1
Historically, testing for gonorrhoea infection at GUM clinics was performed by isolation of N. gonorrhoeae, a specialised technique that requires considerable laboratory expertise. Currently, detection of gonorrhoea is routinely performed using nucleic acid amplification tests (NAATs), which are generally more sensitive and less labour intensive than culture. In the UK NAAT is the test of choice for testing asymptomatic individuals with genital gonococcal infection, and for testing rectal and pharyngeal sites in MSM.2 Current guidelines also recommend that positive NAATs from low prevalence populations and extragenital sites require confirmation by supplementary testing using a different nucleic acid target.2 Despite the advantages and increasing usage of NAATs, culture is required to provide a viable organism for antimicrobial susceptibility testing and to help identify resistant strains of N. gonorrhoeae.
In England and Wales, trends and drifts in susceptibility to antimicrobials in gonoccocal isolates are monitored through the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP). GRASP is an enhanced sentinel surveillance programme that collects data on gonococcal isolates from consecutive patients attending select GUM clinics over a 3-month period every year. This programme was established in 2000 and, in 2012, collected data from a network of 23 laboratories and 25 GUM clinics.3
Regular review of GRASP data is essential to guiding antimicrobial prescribing practice. Since 2004, antimicrobial susceptibility data from GRASP has played a vital role in changes to the national guidelines for treatment of gonorrhoea from ciprofloxacin to cefixime in 2004,4 then from cefixime to ceftriaxone with azithromycin in 2011.5 ,6 Isolates infecting MSM were more likely to exhibit resistance to ciprofloxacin and decreased susceptibility (a minimum inhibitory concentration of ≥0.125 mg/L) to cefixime; representative monitoring of gonococcal isolates is therefore essential to detect any new patterns of antimicrobial resistance. There are multiple reasons why an infection may not be culture-positive, such as the greater sensitivity of NAATs, a long delay between a positive NAAT and performing culture (because the patient received antibiotics or the infection resolved in the interim), or because culture was not offered or was refused by the patient. In this analysis, we sought to determine the representativeness of GRASP isolates by examining the frequency of culture-positivity in patients diagnosed with gonorrhoea in English GRASP clinics, and investigated whether this varied by the clinical, demographic and behavioural characteristics of patients.
Methods
Surveillance system and procedures
Data were obtained from GRASP for the period July–September 2012. The inclusion criteria and procedures for antimicrobial susceptibility testing are described elsewhere.3 ,7 ,8 Briefly, gonoccocal isolates from consecutive patients attending 25 sentinel clinics in England and Wales over the 3-month period were sent to the Sexually Transmitted Bacteria Reference Unit (STBRU), Public Health England (PHE) for antimicrobial susceptibility testing. Depending on clinical history, specimens were collected from the urethra, pharynx or rectum of each patient. Isolates were retrieved, confirmed to be N. gonorrhoeae and the minimum inhibitory concentrations for penicillin, ciprofloxacin, spectinomycin, tetracycline, ceftriaxone, azithromycin and cefixime were determined.3
All GUM patients in GRASP were initially tested for gonorrhoea using NAATs, and demographic and clinical data on all gonorrhoea NAAT-positive patients seen at the 23 English sentinel clinics over the study period were obtained from the routine mandatory surveillance system for STIs at PHE, the GUM Clinic Activity Dataset V.2 (GUMCADv2). These data were linked electronically to the GRASP online reporting form, which enables clinics to enter behavioural (number of sex partners in the UK and abroad in the 3 months preceding diagnosis) and additional clinical (such as the presence of symptoms and site of infection) data. These data were then linked to the records for the gonococcal isolates sent to the STBRU of PHE for antimicrobial susceptibility testing. GUMCADv2 and GRASP data are submitted electronically to the HIV and STI Department of PHE.
Analysis
Data from the first episodes of all gonorrhoea NAAT-positive patients who attended the 23 English GRASP GUM clinics from July to September 2012 were considered in the analysis. During this period, 3103 gonorrhoea diagnoses were reported from the GRASP clinics in England; this comprised 46.1% of all gonorrhoea diagnoses made in England during that period. Of these 3103 diagnoses, 27 (0.9%) were from persons with two distinct episodes of gonorrhoea. To generate a data set containing independent observations, only the first episodes of these 27 patients were retained, leaving a data set of 3076 unique NAAT-positive patients. The outcome for analysis was having a culture-positive N. gonorrhoeae infection, defined as a NAAT-positive specimen for which a gonococcal isolate had been received by STBRU. Patients who did not have culture-positive infections included those who were culture-negative and those on whom culture was not performed. As these groups could not be distinguished, they were treated as a single category in the analysis. For analysis, the number of sex partners was categorised as follows: 0, 1–2, ≥3.
Pearson's χ2 test was used to assess unadjusted associations between patient characteristics and having a culture-positive infection. All variables with significant unadjusted associations were entered into an age-adjusted and clinic-adjusted logistic regression model to determine adjusted ORs (aORs) and 95% CIs, and records were excluded from analysis if data was missing for any variable. A sensitivity analysis was conducted to evaluate the contribution of genital infections to the ‘multiple sites of infection’ variable. All analyses were performed with Stata V.13.1 (StataCorp LP, College Station, Texas, USA), and p values less than 5% were considered statistically significant.
Ethics statement
As GRASP is a routine public health surveillance activity, no specific consent was required from the patients whose specimens were used in this study. Patients were informed about GRASP at every participating clinic using written notices. PHE has permission to handle data obtained by GRASP under section 251 of the UK National Health Service Act of 2006 (previously section 60 of the Health and Social Care Act of 2001), which was renewed annually by the ethics and confidentiality committee of the National Information Governance Board until 2013. Since then the power of approval of public health surveillance activity has been granted directly to PHE.
Results
The median age of the 3076 gonorrhoea NAAT-positive patients was 28 years (minimum–maximum: 13–78 years). Most (67.9%) were white, and 77.2% were born in an European country (62.1% from the UK, 15.1% from another European country). Most (60.1%) were MSM and, in the 3 months prior to diagnosis, 60.9% reported one to two sex partners, and 12.8% reported having sex abroad. Just under half (47.8%) were symptomatic. Based on clinical symptoms, the most commonly reported site of infection was genital (54.3%), followed by pharyngeal (31.4%) or rectal (30.1%). Most (72.6%) of the patients were infected at a single site on the date of attendance, but 20.6% were infected at two or more sites; of these 634 patients with multiple sites of infection, 77%, 75%, 65% and 2% had rectal, pharyngeal, genital or other infections, respectively. Twenty-eight per cent of patients were diagnosed with another STI on the date of attendance, 20.1% had a history of gonorrhoea, and 14.7% were HIV-positive. Of all patients, infection in 1428 (46.4%, 95% CI 44.6% to 48.2%) was culture-positive. There was substantial variability by clinic, with the proportion of culture-positive ranging from 30.5% to 88.7%. The frequencies of select patient characteristics, stratified by gender and male sexual orientation, are given in table 1.
Frequencies of select patient characteristics by gender and male sexual orientation, nucleic acid amplification test (NAAT)-positive* patients (N=3048) at the Gonoccocal Resistance to Antimicrobial Surveillance Programme (GRASP) sentinel genitourinary medicine (GUM) clinics, England, July–September, 2012
The results of bivariate analyses are presented in table 2. Compared with heterosexual men, MSM and women were less likely to have an infection that was culture-positive (44.3% and 39.2%, respectively; p<0.001). Similarly, persons born outside the UK (Europe, excluding UK: 42.6%; non-Europe: 44.0%, p=0.031) and persons with a diagnosis of pharyngeal gonorrhoea (36.8%, p<0.001) or with gonorrhoea at non-genital, non-pharyngeal or non-rectal sites (24.2%, p<0.001) were less likely to have an infection which was culture-positive. Those who were symptomatic (66.5%, p<0.001) and persons with genital infections (62.9%, p<0.001) were more likely to have a culture-positive infection. Those infected at multiple sites (genital, pharyngeal, rectal or other; 61.2%, p<0.001) were also more likely to have a culture-positive infection; this proportion decreased to 55.8% (192/344; results not shown) when genital infections were excluded, but persons with multiple (excluding genital) sites of infection were still significantly more likely to be culture-confirmed (p<0.001). Gender/male sexual orientation, country of birth, being symptomatic and having multiple sites of infection were considered in an age-adjusted and clinic-adjusted multiple logistic regression model; site-specific infection variables were not included due to collinearity with the multiple site infection variable.
Unadjusted associations with culture-positive gonococcal infection, nucleic acid amplification test (NAAT)-positive† patients (N=3076) at the Gonoccocal Resistance to Antimicrobial Surveillance Programme (GRASP) sentinel genitourinary medicine (GUM) clinics, England, July–September, 2012
On multivariable analysis (table 3), infections in MSM and women were less likely to be culture-positive when compared with infections in heterosexual men (aOR (95% CI) 0.74 (0.59 to 0.93) and 0.53 (0.41 to 0.68), respectively). Patients who were symptomatic or presented with multiple sites of infection were more likely to have culture-positive infection (aOR (95% CI) 4.61 (3.93 to 5.42) and 2.15 (1.76 to 2.62), respectively). In the sensitivity analysis considering the revised multiple sites of infection variable (ie, excluding genital), similar (differences between 0.1% and 10.5%) ORs for the variables included in the model in table 3 were observed (results not shown).
Adjusted associations with culture-positive gonococcal infection, nucleic acid amplification test (NAAT)-positive† patients (N=3076) at the Gonoccocal Resistance to Antimicrobial Surveillance Programme (GRASP) sentinel genitourinary medicine (GUM) clinics, England, July–September, 2012
Discussion
In this analysis, gonococcal isolates were available from almost half of all NAAT-positive patients, but with significant differences in the demographic, clinical and behavioural profiles between those who were or were not culture-confirmed. The frequency of culture testing in England and Wales, as well as in other European countries, has probably changed in response to new technologies. Although, in the UK, NAATs offer a more sensitive, cheaper and quicker option for gonorrhoea diagnosis (including supplementary testing for diagnosis confirmation), this convenience comes at the expense of skills for culture.9 Gonorrhoea testing guidance in England currently recommends culture is performed on all patients with a confirmed NAAT result for antibiotic susceptibility testing.10 As new mechanisms of antimicrobial resistance cannot be detected without a viable organism to enable determination of minimum inhibitory concentrations, the likelihood of early detection of emerging antimicrobial resistance is also affected.
A previous study11 has shown that, relative to other clinics in England and Wales, GRASP clinics may over-represent MSM, although weighted analyses correcting this bias had minimal impact on estimates of resistance prevalence. Here we show that, within GRASP clinics, patients whose gonococcal isolates are included in GRASP may not be representative of all gonorrhoea cases. There have been concerns about the emergence of decreased susceptibility to third generation cephalosporins among MSM but, in this analysis, infections in MSM were less likely to be culture-positive; this may be due to an increased level of asymptomatic screening among MSM2 as well as the decreased sensitivity of culture on rectal and pharyngeal specimens.12 In this analysis, the most commonly reported site of infection for the 1845 MSM was rectal (46%) or pharyngeal (46%), followed by genital (38%). Overall, patients with gonorrhoea infection at multiples sites were more likely to be culture-positive and, although this proportion decreased when genital infections were excluded, this had a minimal impact on our results. Symptomatic patients were also more likely to have culture-positive infections. This is unsurprising as clinicians are more likely to request culture from patients with more clinically apparent infection and culture is more likely to be successful from symptomatic patients, which is consistent with national guidelines.6
Data from GRASP suggest that there is considerable variation in susceptibility patterns according to clinical, demographic and behavioural characteristics of those infected. Previously, isolates with decreased susceptibility to cefixime were strongly associated with gender/sexual orientation and ethnicity.5 This highlights the need for representative data from this sentinel surveillance system, as decreased susceptibility to various antimicrobials regularly emerges from distinct population subgroups.
Monitoring of gonococcal antimicrobial surveillance is often restricted to sentinel surveillance. The European Gonococcal Antimicrobial Surveillance Programme collects a predetermined number of consecutive isolates from each contributing country,13 while the Gonococcal Isolate Surveillance Project in the USA similarly collects data from the first 25–30 male patients seen at sentinel clinics each month.14 This might introduce a degree of selection bias, and our analysis also underscores the differences in those who were culture-positive, but the degree to which this affects trends in antimicrobial resistance is unknown. As GRASP is a sentinel surveillance system with data on 46% of all gonorrhoea diagnoses reported during the study period, these findings may not be representative of routine clinical practice and different patients may be more or less likely to have a culture-positive infection outside of the GRASP data collection period. Although we are unable to assess this, it may also be the case that the GRASP clinics, being members of this sentinel network, may be more likely to perform culture than other GUM clinics.
This study also has other limitations. First, the reasons for not having a culture-positive result cannot be distinguished between culture having a poorer sensitivity or because it was never offered to the patient or, if offered, was refused. However, we were still able to evaluate the representativeness of GRASP isolates through comparison of patient characteristics for infections which were or were not culture-positive. Further, there is a high degree of geographical variability in the prevalence of gonorrhoea in England, with the highest prevalence found in urban areas with high degrees of socioeconomic deprivation15 and we know that GRASP over-represents London clinics and infections in MSM.3 ,7 Despite its high sensitivity, the likelihood of a false-positive result on a NAAT increases in areas of low gonorrhoea prevalence, even among GUM clinic attendees,10 thus decreasing the likelihood of having a culture-positive isolate. Additionally, any miscoding in patient identifier codes (∼10% of isolates received in 2012) would prevent linkage of a culture result from the STBRU database to the GUMCADv2/GRASP record, thus artificially decreasing the proportion culture-positive.
GRASP continues to provide vital evidence to review trends in antimicrobial resistance. However, this analysis has demonstrated that under half of gonorrhoea diagnoses made in England have a positive culture for N. gonorrhoeae and that gonococcal isolates collected through GRASP might be biased in favour of infection in specific subgroups of symptomatic patients. This level of coverage may be attributed to the lower sensitivity or lack of consistent offer of culture. Effective patient management and infection control require regular review of treatment guidelines based on reliable and timely evidence. The 2013 GRASP Action Plan16 recommends that laboratories should maintain resources for isolation of N. gonorrhoeae, as it is essential to inform patient management, including investigations of putative treatment failures, and to detect emerging resistance. Furthermore, decreased susceptibility to the cephalosporins requires continued research and development to inform public health interventions. For this to be achieved, routine isolation of N. gonorrhoeae is essential to detect new mechanisms of resistance and ensure representative monitoring of trends in antimicrobial resistance. With case reports of ceftriaxone resistance infection in Europe and Asia,17 ,18 the need to improve capacity for gonococcal culture has also been emphasised in Response and Action Plans by the European Centre for Disease Prevention and Control and WHO, respectively.9 ,19 All primary diagnostic laboratories in England and Wales should aim to develop or continue to maintain skills in culture for N. gonorrhoeae from clinical samples, alongside routine testing using highly sensitive gonorrhoea NAATs, to ensure optimal patient management and to maximise public health benefit.
Key messages
-
Antimicrobial resistance in Neisseria gonorrhoeae is an ongoing concern and testing specimens by culture is necessary to provide a viable organism to monitor this.
-
Less than half of gonorrhoea diagnoses from sentinel surveillance are culture-positive, and these tend to be biased towards specific population subgroups.
-
Routine culture is essential for patient management and surveillance, as well as for representative monitoring of emerging antimicrobial resistance to inform treatment guidelines.
Acknowledgments
The authors thank all GRASP collaborators, including: members of the Sexually Transmitted Bacteria Reference Unit and the HIV and STI Department at Public Health England (J Anderson, N Ellis, M Furegato, B Sile, K Town); the GRASP steering group (D Livermore, C Bignell, K Eastick, A Johnson, C Lowndes, J Paul, A Robinson, J Ross, J Wade); and the collaborating centres—Northampton (M Minassian, L Riddell), Nottingham (V Weston, C Bignell), Cambridge (N Brown, C Carne), Luton (T Balachandran), London Charing Cross, Chelsea and Westminster (K McLean, A McOwan, G Paul), Homerton (A Claxton, M Nathan), Kings (J Wade, M Tenant-Flowers), St George's (P Riley, P Hay), St Mary's (A Smith), University College Hospital (B Macrae, A Robinson, E Jungmann), Woolwich (S Bragman, J Russell), Newcastle (M Valappil, K N Sankar), Liverpool (J Anson, M Bradley), Manchester (A Qamruddin, A Sukthankar), Brighton (M Cubbon, G Dean), Reading (A Tang), Bristol (O M Williams, P Horner), Gloucester (P Moore, A DeBurgh-Thomas), Cardiff (R Howe, R Drayton), Newport (S Majumdar, H Birley), Birmingham (J Gray, J Ross), Wolverhampton (D Dobie, A Tariq), Leeds (M Denton, J Clarke), and Sheffield (C Bowman). We are especially grateful to the collaborating centres that participated in the pilot for the electronic data collection tool for GRASP: Nottingham, Brighton, St. George's, Chelsea and Westminster, University College Hospital and Birmingham.
References
Footnotes
-
Handling editor Jackie A Cassell
-
Contributors GH, CAI, HM, MC and CO designed the study and developed the methodology. NQ performed the laboratory work on GRASP isolates, and SC, MC and CAI supervised the delivery of the microbiological component of GRASP as part of the reference service in STBRU at PHE. CO coordinated the data collection for GRASP and provided the data for analysis. HM wrote the manuscript and performed the analyses. All authors contributed to the revision of the manuscript and have seen and approved the final version submitted for publication.
-
Funding GRASP has been funded totally (2000–2004) and partly (2005–2010) by the Department of Health (England) and by Public Health England.
-
Competing interests None.
-
Ethics approval PHE has permission to handle data obtained by GRASP under section 251 of the UK National Health Service Act of 2006 (previously section 60 of the Health and Social Care Act of 2001), which was renewed annually by the ethics and confidentiality committee of the National Information Governance Board until 2013. Since then the power of approval of public health surveillance activity has been granted directly to PHE.
-
Provenance and peer review Not commissioned; externally peer reviewed.