Article Text

Download PDFPDF

Prevalence of Neisseria gonorrhoeae infection in young subjects attending community clinics in South London
  1. G Gopal Rao,
  2. L Bacon,
  3. J Evans,
  4. Y Dejahang,
  5. P Michalczyk,
  6. N Donaldson,
  7. on behalf of Lewisham Chlamydia and Gonoccoccus Screening Programme
  1. Lewisham Primary Care Trust and University Hospital Lewisham, London, UK
  1. Dr G Gopal Rao, Department of Microbiology, University Hospital Lewisham, London SE13 6LH, UK; gopal.rao{at}uhl.nhs.uk

Abstract

Objectives: To describe the prevalence and epidemiology of gonococcal infection in young subjects attending community clinics in South-East London.

Methods: Subjects <25 years of age participating in the National Chlamydia Screening Programme were tested for gonococcal infection using a nucleic acid amplification test (strand displacement amplification assay).

Results: 10 523 tests were performed in 7369 patients (82% female) over a 2-year period in 2004 and 2005. Specimens used for tests were self-taken vulvovaginal swabs (43%), cervical swabs (40%), urine (16%) and urethral swabs (0.9%). Reasons for tests were: screening (67%), diagnosis (27%) and contacts of patients with chlamydia or gonococcus infection (7%). A significantly higher percentage of male subjects were tested as contacts (19% male vs 4% female; p<0.001). Of the 10 117 cases with definite results, 414 were positive (prevalence 4.1%, 95% CI 3.7% to 4.5%). There was a significantly higher prevalence in male subjects (5.7% male v 3.8% female; p<0.001). The average number of tests was 1.4 per patient (range 1–10). Contacts had a significantly higher prevalence (15.5%, p<0.001) than those tested for diagnostic (3.6%) or screening (3.1%) purposes. Multivariate regression analysis confirmed that there was a significantly higher prevalence in black Caribbean (5.8%, OR 2.44), black British/other black (5.6%, OR 2.33) and mixed (5.5%, OR 2.25) than white (2.4%) ethnic groups (p<0.001). Increasing age was significantly associated with lower prevalence (OR = 0.87; 95% CI 0.84 to 0.91; p<0.001). The odds of a positive test decreased by 13% for every year older.

Conclusion: A community-screening programme has detected a high prevalence of Neisseria gonorrhoeae in South London, especially in teenagers, male subjects and certain ethnic groups.

Statistics from Altmetric.com

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.

The highest rates of gonococcal infection are found in London and urban areas. Within London, North Central London and South London have the highest rates of diagnoses of gonococcal infection.1

The borough of Lewisham is in South London. For the past 2 years, community sexual and reproductive health clinics and a few general practice clinics have been participating in the National Chlamydia Screening Programme (NCSP) in asymptomatic subjects under 25 years of age.2 These clinics also provide diagnostic and contact tracing services for this age group.

In this paper, we describe the prevalence of gonococcal infection using nucleic acid amplification tests (NAATs) in subjects under 25 years of age in South London. To the best of our knowledge, this is the first report of the prevalence of gonococcal infection in young subjects attending community clinics in London.

METHODS (FIG 1)

Study period and location

The study was carried out between January 2004 and December 2005 in community sexual and reproductive health clinics and participating general practices.

Subjects

Subjects under 25 years of age participating in the NCSP also gave informed consent to be tested for gonococcal infection. Patients presenting for diagnostic reasons and contacts of patients with chlamydial or gonococcal infection were also screened as described below.

Specimen collection and transportation

Specimens were either taken by the subject or the clinician. Male subjects were instructed to collect “first pass urines” after holding their urine for at least an hour, and female subjects were given detailed instructions for collection of self-taken vulvovaginal swabs (3 cm into the vagina). High vaginal or cervical swabs were collected if vaginal examination was indicated. The swabs used for NAATs were provided by the manufacturers (BD ProbeTec C trachomatis/N gonorrhoeae Amplified DNA Assay Endocervical and Urethral Collection and Dry transport kits; Becton Dickinson, Oxford, UK).

Subjects with a positive NAAT for N gonorrhoeae were recalled to the clinic as soon as possible. Cervical swabs (Transport swab, Amies medium with charcoal; Sterilin, Stone, Staffordshire, UK) and/or urethral swabs (Transwab, Amies medium with charcoal; Medical Wire and Equipment, Corsham, Wiltshire, UK) were taken before the start of treatment for culture and antibiotic sensitivity tests.

Specimens were transported to the laboratory generally on the same day or stored at 4°C in the clinic and transported at the earliest opportunity.

Laboratory tests

Tests were performed in the microbiology department at University Hospital Lewisham. The NAAT used was based on the strand displacement amplification (SDA) assay (BD ProbeTec, Becton Dickinson). All positive tests were repeated in the laboratory to confirm the results and exclude contamination during testing. A result was reported as positive only if both tests were positive.

To minimise false-positive tests as a result of laboratory contamination, we performed regular (monthly) environmental monitoring for the presence of N gonorrhoeae DNA. No evidence of contamination was found during the entire period of the study. We also excluded staff with upper respiratory infections from the testing laboratory, to avoid contamination with other Neisseria sp.

Culture

Swabs were cultured on selective Gonococcus Medium (Oxoid, Basingstoke, UK) for N gonorrhoeae and incubated in 10% CO2 at 37°C for 48 h. Oxidase-positive colonies were confirmed as N gonorrhoeae by Gram stain, biochemical (Neisseria PET; Bioconnections, Shipley, UK) and immunological (Gonogen II; Bioconnections) tests.

Data retrieval and statistical analysis

Data were retrieved from the laboratory information system (Telepath 2000, iSoft Laboratory Systems, Belper, UK). Prevalence estimations for the overall sample and by subgroup are provided with the corresponding 95% CIs. Univariate comparisons of gonorrhoea and chlamydia positive rates were performed with χ2 tests (Pearson, Fisher exact, or linear trend) for categorical or ordered categorical variables. Mixed multiple logistic regression models were used to assess the independent effect of the different ethnicity groups on the prevalence of gonococcal infection, with adjustment for age and gender, as well as to explore differences between the 2 years of study. The models also adjust for whether a patient was classified as a contact. Interactions between gender and the other factors were tested, and significance was established at the 10% level. Random effects were introduced to account for repeated testing of the same patient. All testing of male subjects was performed on urine or urethral specimens. For female subjects, mainly self-taken vulvovaginal swabs or cervical/vaginal swabs were used. Ethnicity was self-assessed using the chlamydia screening programme collection form.

RESULTS

A total of 10 523 tests for gonococcal infection were carried out in 7369 patients (82% female) over a 2-year period. Reasons for testing were classified as: for screening, 67%; for diagnosis, 27%; on the basis of being reported as a sexual contact, 7%. Of the 7203 tests for screening purposes, 82% were on female subjects. Of the 2811 tests for diagnosis, 95% were on female subjects. Of the 689 subjects tested as contacts, 54% were female. There is a significant association between reason for testing and gender with the percentage of testing, because the patient being identified as a contact was 4% among females and 19% among males (p<0.001).

Of the 10 117 tests with unequivocal results, 414 were positive, giving a prevalence of 4.1% (95% CI 3.7% to 4.5%).

Age and gender

Table 1 shows the distribution of gonococcal infection by age, sex, ethnicity and reason for screening.

Table 1 Distribution of gonococcal infection by age, sex, ethnicity and reason for screening

The prevalence of gonococcal infection in male subjects was significantly higher than in female subjects (5.7% (90/414) vs 3.8% (324/414); p<0.001). Younger female subjects (<16 years) had a higher prevalence than boys in the same age group. Conversely older female subjects (20–24 years) had lower prevalence than men in the same age group. In the 16–19-year age group, the two sexes had similar prevalence rates.

Reason for testing

In female subjects, there was no significant difference in the prevalence of gonococcal infection in those who were tested for diagnostic (3.6%) and screening (3.2%) purposes. However, the prevalence was significantly higher in those tested as contacts (15.5%; p<0.001) in both sexes and for diagnosis in male subjects (15.4%; p<0.001).

In total, 62% of male subjects (56/90) and 45% of female subjects (147/324) with gonococcal infection were tested in order to make a diagnosis or were contacts of patients with chlamydial or gonococcal infection.

Ethnicity differences

The highest prevalence was found in black Caribbean (5.8%), black British/other black (5.6%) and mixed (5.5%) groups. The prevalence in other ethnicities was: unknown/unspecified (4.4%), Asian subcontinent (3.1%), black African (2.7%), white (2.4%), and other ethnic groups (1.7%). These observed differences in the prevalence were significant (χ2 test (8 df) = 71; p<0.001).

Multivariate logistic regression analysis

Table 2 shows the association between demographic characteristics and gonococcal infection. The average number of tests was 1.4 per patient (range 1–10). The odds of a positive test decreased by 13% for each year older (OR = 0.87; 95% CI 0.84 to 0.91; p<0.001). We also observed that the odds of being positive increased by 27% between 2004 and 2005 (OR = 1.27; 95% CI 1.03 to 1.56; p = 0.03). The odds of a positive result for those whose reason for visiting the clinic was “contact” was more than five times as high as for those whose reason was diagnosis and screening (OR = 5.48; 95% CI 4.30 to 6.99, p<0.001). After adjustment for these effects, in relation to the white group, the odds of testing positive were over two times greater for the black British/other black (OR = 2.33, 95% CI 1.74 to 3.13, p<00.1), black Caribbean (OR = 2.44, 95% CI 1.82 to 3.27, p<0.001) and mixed (OR = 2.25, 95% CI 1.55 to 3.25, p <0.001) groups. The odds were almost twice as high in relation to white subjects for the unknown/unspecified group (OR = 1.94, 95% CI 1.01 to 3.72, p = 0.05). None of the interactions between gender and the other factors were found to be significant.

Table 2 Multivariate logistic regression analysis: association between demographic characteristics and gonococcal infection

Co-infection with Chlamydia trachomatis

Co-infection with C trachomatis was detected in 55% of the patients with N gonorrhoeae infection (222/406).

Correlation between positive NAAT and culture for N gonorrhoeae

We received 296 (72.9%) specimens for culture from 406 subjects who were found to be positive by NAAT. The most common reasons for culture results not being available were: refusal for swabs to be taken; non-attendance at clinic for culture; treated in a genitourinary medicine (GUM) clinic. About two-thirds (199/296; 67.2%) of the specimens yielded N gonorrhoeae.

DISCUSSION

NCSP has established the importance of screening asymptomatic young subjects in various urban communities in England. The programme has detected high prevalence (10–15%) of asymptomatic chlamydial infection.2 However, in the absence of a national gonococcus screening programme, little is known about the prevalence of gonococcal infection in young subjects in the community. As seen in the NCSP, confining screening for gonococcal infection to GUM clinics only may seriously underestimate the burden of infection in the community.

NAATs amplify and detect DNA specific for N gonorrhoeae. A principal concern has been the false-positive rate with some of the earlier NAATs and the consequent clinical and medicolegal implications.3 A recent systematic review of NAATs concluded that PCR, transcription-mediated amplification and SDA assays are sufficiently sensitive to be used for N gonorrhoeae screening. However, the sensitivity of PCR in female subjects is too low to recommend its routine use to test for gonorrhoea in urine specimens.4 Other reports also confirm that SDA assay has high specificity and sensitivity.5 The high specificity and sensitivity recorded by these studies using culture as a gold standard suggests that a positive NAAT is indicative of true infection capable of causing symptoms and transmission. It has also been shown that NAATs for N gonorrhoeae become negative within 2 weeks of successful treatment.6

Key messages

  • This is the first report of a community-based screening programme using nucleic acid amplification tests for gonococcal infection in South London in subjects under 25 years of age.

  • A higher rate of gonococcal infection was seen in the community than in those attending genitourinary medicine clinics.

  • Infections were significantly more common in younger patients, males and certain ethnic groups.

It has been shown that transcription-mediated amplification tests have a positive predictive value of at least 95% in female subjects in settings where prevalence of gonococcal infection is as low as 0.5%.7 This indicates that false-positive tests with NAAT are rare even in low-prevalence populations.

In this study, we were able to confirm the positive NAAT results by culture in only 67.2% of the subjects. In an audit of a similar screening programme in Lambeth (a neighbouring borough of Lewisham), whereas 85% of positive NAAT results using SDA assay results were confirmed by culture in patients attending the sexually transmitted infection (STI) clinic in the hospital, only 55% of the positive results were confirmed when the culture specimens were taken in clinics in the community (personal communication, J Klien, Consultant Microbiologist, St Thomas’ Hospital, London). This suggests that our culture results are likely to be false-negative rather than NAAT being false-positive. The relatively low confirmation rate in the community may be due to survival characteristics of different auxotypes of N gonorrhoeae or suboptimal collection, storage and transport.8 9 Since the introduction of improved specimen collection and transportation procedures, we have observed an improvement in culture confirmation in our laboratory. Currently we are able to confirm the 80–85% of positive NAATs with culture (unpublished data). We are also undertaking a study to see if the culture positive rates can be improved by “on site” inoculation and incubation of culture plates, as is the practice in many GUM clinics.

Prevalence of gonococcal infection in Lewisham

In this study, the overall prevalence of gonococcal infection was 4.1%. This rate is higher than the highest rate detected in subjects attending GUM clinics in North Central London (1.17%) and South London (1.04%) in 2004.1 In female subjects, the prevalence was 3.8%. Female subjects tested for screening and diagnostic purposes had similar rates, suggesting that it may be difficult to differentiate between these two categories. Although the prevalence of infection was higher in male subjects (5.7%), the latter constituted only 16% of all the patients tested. The high rate of infection in the young men tested may reflect the fact that they were more likely to present once they are known to be contacts of a patient with STI. The Health Protection Agency’s London Surveillance Bulletin recorded that there has been a gradual rise in laboratory-confirmed gonococcal infection in young males in London in 2006.10 For the same period, no increase in gonococcal infections were noted in the GUM clinics.

Female patients accounted for 78% of the cases, and 55% of female and 38% of male patients with gonococcal infection were detected by opportunistic screening. Our findings indicate that asymptomatic infection is common in both sexes. The natural history of asymptomatic infection is unclear, but it is likely to play an important role in the transmission of gonococcal infection in the community.11 12

The implications of treatment of asymptomatic gonococcal infections in prevention of transmission, development of symptoms or future complications are unclear. It has been debated whether all high-risk subjects should be screened for gonococcal infection or only those who have chlamydial infection. A recent study in the Netherlands13 reported that nearly all subjects who had gonococcal infection also had chlamydial infection, and recommended that only subjects with chlamydial infection should be tested for gonococcal infection. However, in our study, 45% of the subjects with gonococcal infection did not have chlamydial infection. Our findings show that selective screening for gonococcal infection in subjects with chlamydial infection is likely to miss less than half of the cases in a community-based screening programme. These findings confirm those of Creighton et al,14 who studied co-infections with C trachomatis and N gonorrhoeae in a GUM clinic in South London. They found that 24.2% (124/512) of heterosexual male patients and 38.5% (136/353) of female patients with gonorrhoea also had chlamydia (p<0.001). Conversely, many of the heterosexual male and female patients had gonococcal infection without chlamydial infection.

Demographics of infected subjects

Younger patients, males and certain ethnic groups were more likely to be positive. These findings are similar to those of the NCSP.1 Female subjects in the <16-year and 16–19-year age groups and male subjects aged 16–24 were at the greatest risk of infection. Our study found that the prevalence of infection was significantly higher in black Caribbean, black British/other black and mixed ethnic groups than in white ethnic groups. These findings confirm those of Low et al15 in a local GUM clinic.

CONCLUSIONS

Subjects under 25 years of age in South London attending community clinics were screened for gonococcal infection using SDA assay. The prevalence of gonococcal infection was high (4.1%) particularly in the young, males and certain ethnic groups.

Although a screening programme for gonococcal infection in conjunction with NCSP is likely to meet most of the criteria set by the national screening committee, its cost effectiveness will depend on the prevalence of gonococcal infection in the local population.1618

Acknowledgments

We thank Professor Catherine Ison for her helpful suggestions.

Lewisham Chlamydia and Gonococcus Screening Group consists of: GGR, LB, YD, Natasha Espinosa, Rosie Jackson, James Wong, Ruth Hardwick, PM, JE, Ruth Williams, Hazel Fisher, Jane Miller, Charles Mazhude, ND.

GGR is the Chairman of the Lewisham Gonococcus Screening Group and contributed to the writing of the paper and is the guarantor. LB is the overall head of the STI programme and contributed to the planning of the programme and the writing of the paper. JE is the clinical lead for the STI programme and contributed to the planning of the programme. YD is the nurse coordinator of the programme, contributed to extraction and review of the data from the clinical database, and contributed to the writing of the paper. PM contributed to extraction and synthesis of the data from the clinical and laboratory databases, and contributed to the writing of the paper. ND analysed the data and contributed to the writing of the paper.

REFERENCES

Footnotes

  • Competing interests: None declared.

Linked Articles

  • Brief encounters
    Helen Ward