In the United Kingdom, infections caused by Neisseria gonorrhoeae are predominantly diagnosed and treated free of charge in open-access genitourinary medicine (GUM) clinics. National statistics on gonorrhoea and other sexually transmitted infections (STIs) diagnosed in these GUM clinics are collated by the Health Protection Agency (HPA) in aggregate form via KC60 returns.1 Gonorrhoea diagnoses fell sharply in the 1980s and early 1990s, largely owing to safer sexual behaviours brought about in response to the HIV epidemic.2 Diagnoses began to rise after 1994, peaked in 2002 and have fallen subsequently. Gonorrhoea remains concentrated in large urban areas, particularly London, with a rate of 157/100 000 population. Rates among men outnumber those among women because of a high number of diagnoses in men who have sex with men (MSM) and because infections in women are more likely to be asymptomatic.1

Detailed demographic, behavioural and risk information is critical to understanding the epidemiology, informing public health policy and to evaluating the success or failure of health promotion activities aimed at reducing the transmission of STIs. While the KC60 returns provide useful national trend data on STIs diagnosed at GUM clinics, they contain limited information regarding the risk behaviour of the individuals who acquire infections. To gain a better insight into the epidemiology of gonorrhoea in England and Wales, we analysed the enhanced dataset collected through the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP)3 for the years 2001–6. Established in 2000, GRASP primarily monitors trends in antimicrobial resistance among gonococcal isolates to inform treatment guidelines,4 5 but also provides detailed epidemiological and behavioural profiles of individuals diagnosed with gonorrhoea at sentinel GUM clinics in England and Wales.

METHODS

Sites

Twenty-six sentinel GUM clinics and their supporting laboratories in England and Wales take part in the programme each year, nine in London and 17 elsewhere. The clinics outside London were originally selected on the basis of geography, two per health region, and the burden of disease, having greater than 100 diagnoses per annum. The health regions were however changed in 2002, and the new North East region has only one GRASP site.

Ethical permission was obtained from local regional research committees and from the North West multicentre research ethics committee. Patients were informed of the study at the participating site through written notices.

Population and analyses

All patients clinically diagnosed with gonorrhoea and with a corresponding isolate of Neisseria gonorrhoeae obtained at a participating laboratory between the months of June and August for the years 2001–6, were included. Only one isolate of N gonorrhoeae per patient was tested and where multiple sites were infected an order of preference was assigned (in order: men: rectal, urethral, any other site; women: cervical, any other site).

Data were collected using a standard form and included gender, age, sexuality, ethnicity, concurrent STIs at attendance, symptoms, site of infection, previous episodes of gonorrhoea, number of sexual partners over the previous three months within the United Kingdom and abroad and therapy prescribed. HIV status has been collected since 2004. Duplicate cases were removed from the dataset if the same patient presented more than once within a 28-day period. Tests of significance are either by Pearson’s χ² test or by Fisher’s exact test. Proportions were calculated, taking all cases with available information as the denominator, but excluding cases with missing values.

RESULTS

Trends

A total of 12 282 individuals were reported to GRASP over the period (2001 to 2006) with just under half (47%) of diagnoses reported from London sites. The number of cases reported annually declined from 2513 in 2001 to 1599 in 2006, a 36% decrease. The proportion of diagnoses among men who have sex with men (MSM) significantly increased from 25% in 2001 to 39% in 2006 (p<0.005).

Epidemiological and clinical characteristics

Almost three-quarters (73%) of cases were men, of whom 40% (3310/8174) were MSM. Women were younger, with median age of 20 years at diagnosis compared with 25 and 31 years among heterosexual men and MSM, respectively.

Diagnoses varied by gender, sexual exposure (men) and ethnicity, with women and MSM most likely to be white, and heterosexual men most likely to be black (table 1). These aspects however changed over time and, among heterosexuals, there was a striking decline in the proportion of diagnoses among those of black Caribbean ethnicity, declining from 43% (444) in 2001 to 30% (142) in 2006 in men and from 205 (32%) to 68 (21%) in women.

The majority of cases were infected at one anatomical site (table 1) with the cervix the most common site of gonococcal isolation in women (98%) and the urethra in both heterosexual men (98%) and MSM (65%). Rectal and pharyngeal infections were reported in 35% and 15% of MSM, respectively.

Heterosexual men were most likely to present with symptoms, followed by MSM and then women (table 1). A history of gonorrhoea was common in all groups, but most common in MSM (table 1). Women were more likely to have a concurrent STI, excluding HIV, diagnosed at the same time as gonorrhoea, than men (50% compared with 27%, p<0.0005). Chlamydia was the most common concurrent pathogen across all groups (table 1).

Sex abroad was a relatively small but significant risk factor (table 1). Among cases with this risk factor, the majority of MSM (65%) indicated a partner in another western European country, whereas 57% of the women reported contact in the Caribbean and heterosexual men reported partners in a variety of regions including the Far East (25%), Caribbean (21%) and another western European country (21%).

The average number of sexual partners reported over the previous three months was 1.7 for women, 2.2 for heterosexual men and 4.7 for MSM, with a median of one partner for women and two for both heterosexual men and MSM. The majority of women reported a single sexual partner in the previous 3 months, whereas more than half of heterosexual men and MSM reported two or more sexual partners (table 1). Significantly more MSM than heterosexual men reported six or more sexual partners (18% of MSM vs 4% of heterosexual men; p<0.0005).

HIV status was available for the years 2004–6. Where available, a positive HIV status was recorded for 31% of MSM. HIV-co-infected MSM were older at presentation than HIV-negative men (median 35 years vs 28 years), more likely to present at a London site (70% vs 52%, p<0.0005), to have a history of gonorrhoea (66% vs 36%, p<0.0005) and to have a concurrent STI (28% vs 20%, p = 0.002). HIV-positive MSM had a higher average number of sexual partners in the three months before their gonorrhoea diagnosis (6.8 vs 4.3 for HIV-negative men).

Patients aged <25 years compared with those aged >25 years

Seventy-three per cent of the infected women, 47% of the heterosexual men and 22% of the MSM were under 25 years. The younger women were more likely to have a concurrent STI than older women (55% vs 36%, p<0.0005), and despite a probable shorter sexual history, given their younger age, up to 20% reported a history of gonorrhoea, compared with 22% for older women. Similarly, 37% of young heterosexual men presented with a concurrent STI and a quarter (25%) had a history of gonorrhoea, compared to 27% and 33% in older men, respectively. The same proportion of young and older MSM had a concurrent STI (21%). About a quarter (28%) of young MSM reported a history of gonorrhoea compared with 45% in older MSM. Numbers of sexual partners reported in the previous three months did not differ by age group in women or men.

DISCUSSION

GRASP has provided a detailed description of the epidemiology of gonorrhoea in England and Wales and its findings show that many of the infected patients are repeatedly engaging in unsafe sexual behaviours. In particular, the GRASP data confirm the high burden of gonococcal infection in core groups of MSM, particularly those that are HIV positive, and young heterosexual adults, particularly those with a history of gonorrhoea and those of black Caribbean ethnicity.

Most of the MSM were of white ethnicity, predominantly aged between 25–34 years and report multiple sexual partners in the previous three months. One in seven (14%) probably acquired their infection abroad. The high proportion (31%) of HIV co-infection is alarming, especially since HIV-positive men were more likely to report previous gonorrhoea and concurrent STI and had a higher average number of sexual partners than HIV-negative men. These findings indicate that HIV-positive men are at higher risk of acquiring further STIs than HIV-negative men. Selection of sexual partners, based on known HIV status (termed “serosorting”), and choosing not to use a condom are likely to increase the risk of other STIs. High rates of gonococcal rectal infections (35%) among MSM provide evidence of unprotected receptive anal sex and are of concern, for both HIV positive and negative men. These data are supported by behavioural surveys showing continued high-risk sexual behaviours and inconsistent condom use among MSM, including those that are HIV positive.5^–9 Consequently MSM remain the group most at risk of STIs in the UK1 10 11 and internationally.12^–14 There is an urgent need to strengthen prevention efforts in gay men, ensuring targeted interventions for both HIV positive and negative men across all age groups.

Men who acquired gonorrhoea through heterosexual contact were mostly aged between 20 and 34 years; overall one in three was of black Caribbean ethnicity, though this proportion fell over time. Like MSM, most heterosexual men reported multiple sexual partners in the three months before diagnosis and about one in 10 probably acquired their infection while abroad. There were marked differences in the characteristics of infected women compared with both heterosexual and homosexual men. Unlike heterosexual men, most women are diagnosed as young adults (three-quarters are aged less than 25 years) and most (68%) report a single sexual partner in the previous 3 months. A disproportionate number were of black-Caribbean ethnicity or of another black ethnicity. The lower median age at diagnosis of women is likely to reflect patterns of sexual mixing within the heterosexual population.15

High re-infection rates (one in five had a previous episode of gonorrhoea) and concurrent infections with another STI (especially chlamydia) is of particular concern across all groups and underscores the need for prompt partner notification and treatment. Furthermore, over half of women were asymptomatic for gonorrhoea indicating a potential reservoir of undiagnosed gonococcal infections. Screening for gonococcal infection performed in young people participating in the National Chlamydia Screening Programme in England16 in areas of London suggest a high rate of gonorrhoea among those screened.17 Close monitoring and evaluation of such initiatives, particularly in view of possible limitations of nucleic acid amplification tests (NAATS), are required to inform future guidance in this area.

The over-representation of diagnoses among black Caribbean men and women who acquired their infection heterosexually is striking, given that people of black Caribbean ethnicity comprise only 1% of the population of England and Wales.18 Local studies have previously identified ethnic minorities as being particularly at risk of gonorrhoea within the inner London area.19^–21 Our findings confirm similar disparities across the country.22 23 Racial disparities in the incidence of bacterial STIs reported in the United States have been attributed to poor access to quality health services and racial differences in sexual mixing and network patterns.24 25 Poor access to healthcare is less likely to play a part in the UK given the availability of universal and free access to sexual health services. Furthermore there have been substantial improvements in access to GUM clinics since it was identified as a major priority area in the National Strategy for Sexual Health and HIV government (2001–2010) for England.26 HPA data indicate a rise in the overall percentage of patients seen within 48 hours at GUM clinics from 48% in 2004 to 72% in 2007 (it was 65% in November 2006 at the end of the study period).27 Although no ethnicity data are available, the survey was conducted on all patients attending all GUM services over a one-week period and is therefore believed to be representative of all STI attendees. Rather, a high background prevalence of gonorrhoea, relatively infrequent inter-racial sexual mixing, high partner change (men) and frequent travel to or migration to the Caribbean are likely to be contributing factors.

Recent declines in gonorrhoea diagnoses among heterosexuals, and particularly those of black-African ethnicity, are encouraging and reasons for this trend should be explored further. It is possible that improvements in partner notification and a decrease in GUM clinic waiting times have contributed to this decline.28 The failure of these initiatives to reduce infection rates among MSM may reflect their higher partner change rate.22 28 The decline in gonorrhoea diagnoses in heterosexuals reported in our study is reflected in national statistics. This is in contrast to diagnoses of other STIs, such as chlamydia, genital herpes, warts and HIV, which continue to rise.1 The often complex clinical presentation of these STIs is likely to make their detection, diagnosis and treatment more challenging than for gonorrhoeal infections.

The decline in gonorrhoea diagnoses in heterosexuals reported in our study is reflected in national statistics. This is in contrast to diagnoses of other STIs, such as chlamydia, genital herpes, warts and HIV, which continue to rise.1 The often complex clinical presentation of these STIs is likely to make their detection, diagnosis and treatment more challenging than for gonorrhoeal infections.

There are several limitations to our study. The 26 sentinel sites participating in the GRASP programme were selected on the basis of high throughput to optimise regional coverage and may not be entirely representative. Furthermore, GRASP primarily collects data from GUM clinics and will therefore miss patients diagnosed in primary care and other settings, although laboratory data indicate that there are few diagnoses is these settings.1 Nevertheless monitoring of gonorrhoea and other STIs in these settings will become increasingly important given shifts towards greater testing of STIs outside of GUM clinics. Greater use of non-culture-based diagnostic methods involving nucleic acid amplification is likely to impact on this trend.

The findings of this study highlight the highly concentrated nature of gonorrhoea presenting at GUM clinics within specific at-risk groups, many of whom have reported previous STI infections. This underscores the importance of partner notification and the urgent need for targeted interventions of proven effectiveness.