There have been substantial increases in notifications of gonorrhoea in men who have sex with men (MSM) in several countries over the last decade.1,2 The rates will likely continue to increase with wider use of biomedical HIV interventions, such as preexposure prophylaxis or treatment as prevention which have been associated with increases in the number of sexual partners and reductions in condom use.3,4 Increasing rates of gonorrhoea are likely to be associated with parallel increases in the rates of antibiotic resistance in Neisseria gonorrhoeae (NG) isolates.5 In response to this resistance threat, the World Health Organization and US Centers for Disease Control and Prevention have recommended a focus on reducing the prevalence of gonorrhoea as a key strategy to reduce the increase in cephalosporin-resistant NG.
Reducing the prevalence of gonorrhoea in MSM requires a thorough understanding of what drives transmission. Existing epidemiological data suggest the oropharynx plays a central role in transmission among MSM. Because of unique behavioral characteristics, asymptomatic sites of infection are principal drivers of gonorrhoea prevalence in MSM.6 Two prospective cohort studies showed that the incidence of oropharyngeal gonorrhoea was substantially greater than the incidence of either rectal or urethral gonorrhoea in MSM.7,8 Both studies are likely to have underestimated the true incidence of oropharyngeal infection because oropharyngeal gonorrhoea is likely to resolve spontaneously in less than 3 months,9 which was less than the average screening interval in both studies.
Epidemiological studies have also implicated the oropharynx, with studies showing that one third of rectal gonorrhoea infections occurred in the absence of unprotected anal sex, and one third of urethral gonorrhoea was attributable to oral sex.8,10 A recent study has shown that if an oropharyngeal swab is positive for NG then invariably so is the saliva sample from the same person.11 About 69% of MSM attending an Australian sexually transmitted infection service used their partner’s saliva as a lubricant for anal sex and fingering in the past 3 months,12 a finding similar to that of a large US study.13 It is important to acknowledge that separating the role of saliva is particularly difficult because it is involved in kissing, oral sex, and anal sex if oral-anal contact occurs before penetration, regardless of condom use. Given the potential role of saliva in gonorrhoea transmission, it is possible that the use of daily mouthwash could reduce NG transmission, with recently published data supporting this intervention.14 We hypothesized that the rate of gonorrhoea is much higher in MSM than heterosexuals because of unique behavioral characteristics of MSM that allow asymptomatic infection at the oropharynx to be the principle driver of the prevalence.6
In the absence of extensive empirical data, mathematical models provide a powerful method for investigating the role that the oropharynx or saliva may play in NG transmission among MSM. Models can also be used to investigate the effectiveness of hypothetical interventions to reduce the prevalence of infection. To date, only one model that accommodated site-specific NG transmission has been published, but it did not include kissing as a potential route of transmission and was not calibrated to the latest data.15 We developed a mathematical model of gonorrhoea transmission that aimed to explore several key factors that underpin the transmission of NG among MSM in Australia. We also used the model to investigate the potential population impact of using daily mouthwash to reduce the duration of oropharyngeal gonorrhoea infection. We also compared these outcomes with those that could be achieved by scaling-up oropharyngeal and rectal gonorrhoea screening, as recommended by current sexually transmitted infection guidelines.16
Disease Burden, Clearance of Gonorrhoea, and MSM Risk Behaviors
We used diagnosis data of gonorrhoea infection at the oropharynx, urethra, and rectum among MSM attending Melbourne Sexual Health Centre (MSHC),17 as there are no Australian population-based prevalence data available from the last 10 years. For the prevalence of gonorrhoea at the oropharynx and rectum, we used estimates of the proportion of MSHC MSM attendees who were diagnosed with the infection by nucleic acid amplification testing (NAAT). Oropharyngeal and rectal prevalence was found to be 8.6% (95% confidence interval [CI], 7.7–9.5%) and 8.3% (95% CI, 7.4–9.1%), respectively, from 7415 MSM attendees in 2015. All these attendees, regardless of the presence of symptoms, were screened routinely for gonorrhoea infection at all 3 anatomical sites when they were presented to the clinic. Notably, these prevalence levels were substantially higher than previous findings that were tested with bacterial culture at the same study site.17 Approximately 2.3% of MSM who attended a sexual health service in a year were found to have symptomatic urethral gonorrhoea infection and actively sought treatment.17 Because most cases of urethral infection present shortly after symptoms appear and are treated, they will resolve on average within 1 week.18 This indicates that the prevalence of urethral gonorrhoea infection in the entire MSM community at a given point in time will be much lower than 2.3%. Asymptomatic urethral gonorrhoea is uncommon (10%), but when it occurs, it is (19,20) likely to remain infectious for 3 to 5 months before natural clearance. Therefore, the proportion of urethral gonorrhoea cases that are potentially infectious will be 2.3% multiplied by 1/52 (infectious for one week till treatment) plus an additional asymptomatic 10% of cases (ie, 0.23%) who will be infectious for 3 to 5 months. Based on this information, we estimated the prevalence of urethral gonorrhoea to be about 0.20% (0.04–0.35%) in the MSM community Supplemental Digital Content 1–2, (Table), https://links.lww.com/OLQ/A203.
Previous literature has documented that behavioral patterns associated with high-risk core groups is essential for sustaining transmission of NG in general populations.21 Based on the STIGMA guidelines, we defined “high-risk” MSM as those having more than 10 partners in the past 6 months, and the low-risk group as those who do not.16 This leads to an estimate of 18% (12–24%) of “high-risk” MSM from published literature.22,23 Based on our clinical sample, oropharyngeal, rectal, and urethral prevalence in high-risk MSM were found to be 10.6% (8.9–12.3), 9.2% (7.6–10.8%), and 0.25% (0.06–0.45%), respectively, whereas the corresponding prevalence in low-risk MSM were 8.1% (7.4–8.7%), 8.0% (7.3–8.6%), and 0.21% (0.07–0.35%) (Table S2).
We conducted a literature review of sexual behavior data for Australian MSM in the period 2004 to 2014 (Table S1). When untreated, oropharyngeal and rectal gonorrhoea infections persist for an average of 12 and 50 weeks,9 respectively. The duration of untreated urethral gonorrhoea was assumed to be similar to that of oropharyngeal infection,15 but is reduced to a week when treated.18 The duration of untreated oropharyngeal gonorrhoea of 12 weeks corresponds to a natural clearance rate of approximately 1% per day. The latest empirical data from a randomized controlled trial and in vitro experiments suggest that mouthwash can substantially increase the clearance rate of NG at oropharynx.6,24 Currently, about 40% MSM received site-specific (all 3 sites) gonorrhoea screening annually,22 which is included in the baseline scenario of the model.
We constructed a model (Fig. 1) to simulate the transmission of NG in MSM. The model accounted for the heterogeneity due to sexual mixing of high-risk and low-risk subgroups in MSM and was informed by epidemiological and behavioral data collected for these 2 subgroups separately. A sexual mixing matrix represents the pattern of sexual “assortativity” and interaction between risk subgroups. A “mixing index” (α) denotes the level of mixing and is often represented as the ratio of the sum of the off-diagonal entries in the mixing matrix and the sum of diagonal entries. The mixing index approach to zero when there is no sexual mixing between the subgroups and increases to an asymptotic value when the subgroups are ‘well-mixed’. We used an assortativity factor α = 0.5 in our analysis.
The model employed a susceptible-infected-susceptible (SIS) compartmental structure expressed as a system of ordinary differential equations (Appendix). At each anatomical site (oropharynx, urethra, rectum), susceptible individuals become infected through a ‘force of infection’ λ and clear infection at a rate γ. The system of differential equations is said to be at equilibrium when NG prevalence at the oropharynx (Po), rectum (Pr), and urethra (Pu) all reach a steady state.
Per-Act Transmission Probability for Gonorrhoea
We estimated transmission probabilities for all modes of transmission using Monte-Carlo simulations.25 This includes 7 rates: 2-way transmission between oropharynx and urethra during oral sex, 2-way transmission between urethra and rectum during anal sex, 2-way transmission between oropharynx and rectum during rimming, and transmission from oropharynx to oropharynx during kissing. Transmission probability from rectum to oropharynx is lower than that from rectum to urethra, because the penis is more likely to make contact with the site of infection during anal sex when condoms are not used (Fig. 1). For each simulation, the rates of transmission were randomly sampled from the interval [0 1] using Latin hypercube sampling, and all biological, epidemiological and behavioral parameters were also sampled with the same approach from their corresponding CIs (Table S1). The set of sampled indicators were input to the system of differential equations to generate a theoretical equilibrium prevalence, which was then calibrated to the corresponding empirical prevalence at each anatomical site in both risk subgroups and also the prevalence in the overall population. The difference between the theoretical and empirical data was measured by the sum of the least squares. This process was repeated for 10,000 times, and the top 1% best-fit simulations were used to obtain the best estimate and CIs of the transmission probabilities.
Assessing the Population Impact of Mouthwash
The population impacts of the potential use of mouthwash in Australian MSM depend on the perceived efficacy of the mouthwash, the proportion of MSM using the product and the frequency of use. We constructed simulation scenarios for potential efficacies of mouthwash that would additionally reduce 0.5%, 1.0%, and 1.5% of the infection duration for each episode of mouthwash use. For each efficacy value, we assessed changes in prevalence and incidence of gonorrhoea as a function of the combined effects of coverage and frequency of use, compared with the baseline scenario of no mouthwash use. The incidence of gonorrhoea was defined as the cumulative number of new cases per year and was estimated as the product of the proportion of the susceptible population and the force of infection.
Data and Model Uncertainties
All sampled model parameters were included with uncertainty bounds to account for their effects on the estimates of transmission rates, and prevalence and incidence levels of gonorrhoea (presented with median values and 95% CIs).
Based on the clinical data at MSHC, we estimated prevalence of oropharyngeal, rectal and urethral gonorrhoea to be 8.6% (7.7–9.5%), 8.3% (7.4–9.1%), and 0.20% (0.04%–0.35%). The model-estimated per-act transmission probabilities for NG was the highest for transmission from urethra to rectum (0.458 [0.210–0.894]) and from urethra to oropharynx (0.403 [0.204–0.777]) (Fig. 1). Notably, the probability of oropharynx to oropharynx transmission through kissing was 0.172 (0.023–0.206) per act.
Oropharynx to oropharynx transmission of NG accounts for the majority of new cases in all scenarios. In the absence of any interventions (Fig. 2), it is estimated that there would be 33.2 (25.0–41.3) incident cases per 100 MSM per annum among which, oropharynx to oropharynx transmission through kissing accounts for nearly three quarters of the cases (71.6% [64.4–80.5%]).
Figure 3 shows the effect of different screening frequencies on site-specific prevalence of infection including the current annual level of about 40%.22 Our analysis shows that increasing oropharyngeal screening to 100% alone will reduce oropharyngeal, rectal, and urethral prevalence by approximately 40% to 5.6% (4.9–7.1%), 5.6% (4.6–7.3%), and 0.15% (0.14–0.16%), respectively (Fig. 3). By contrast, 100% rectal screening alone does not have a significant impact on oropharyngeal infection prevalence, but reduces rectal and urethral prevalence to 4.8% (4.6–5.4%) and 0.15% (0.14–0.19%), respectively. Further, 100% coverage for both rectal and oropharyngeal screening will reduce prevalence of oropharyngeal and rectal infections by about 50% to 5.0% (4.7–5.8%) and 3.2% (3.0–3.9%), respectively. In comparison, daily use of mouthwash has a major impact in reducing the prevalence of gonorrhoea among MSM (Figs. 3–4). A low effective mouthwash (increases daily clearance by an additional 0.5%), together with a moderate coverage (50%) and frequency of usage (50%), would reduce oropharyngeal, rectal, and urethral infection to 5.2% (4.4–6.6%), 5.2% (4.3–5.9%), and 0.14% (0.13–0.15%), respectively. A moderate efficacy of 1% reduces prevalence substantially, and an efficacy of 1.5% efficacy would reduce the percentage of infected cases below 5% (Fig. 3).
We investigated the impact of our assumptions regarding the proportion of asymptomatic infection at the urethra and the duration of untreated oropharyngeal infection on the model findings. In the unlikely scenario where 50% of urethral infection is asymptomatic, mouthwash became less effective, but a moderately/highly effective mouthwash may still reduce gonorrhoea prevalence at all sites by 50% to 75%. Halving the duration of untreated oropharyngeal infection to 6 weeks would lead to only approximately 40% to 55% reductions in prevalence for a moderately/highly effective mouthwash. Doubling the duration of untreated symptomatic urethral infection changed little to the results regarding the population impacts of mouthwash.
Our findings are consistent with oropharyngeal infection playing a key role in sustaining gonorrhoea among MSM and that kissing alone may contribute up to about 75% of incident cases. It predicts that even 100% coverage of oropharyngeal and rectal screening annually will have only a moderate effect on reducing gonorrhoea prevalence, which is consistent with the rising gonorrhoea rates despite increases in screening.2,17 In contrast, frequent use of mouthwash may be more effective at reducing gonorrhoea in MSM. Our model suggests that if mouthwash were effective and widely used, rates of gonorrhoea in MSM may fall precipitously. Given the urgency and potential benefits of this intervention, we suggest empirical research should now be undertaken to test these model findings.
Our model suggests that kissing has a dominant mode of NG transmission in MSM. Indeed, our model could not replicate the current prevalence data without including this mode of transmission. This finding is supported by the observation that saliva frequently has NG detected in it. In a recent study, all 31 MSM who tested positive for oropharyngeal NG by NAAT, also tested positive in saliva.11 The observation that gonorrhoea is substantially more common in younger MSM, who also kiss considerably more than older MSM also supports kissing as a transmission mode.6,26 In contrast, chlamydia, which is not transmitted commonly from the oropharynx is not more common in younger MSM.17 The arguments supporting kissing as a common route of transmission are discussed in detail in another article.6
Despite a high transmission probability (~50%) from the urethra to the throat or rectum, the proportion of incident cases attributable to transmission from the urethra is quite low (urethra to rectum, 1.0%; urethra to oropharynx, 2.3%, respectively). Transmission from the urethra is low because the duration of urethral infections is very short due to rapid appearance of symptoms and hence treatment. Therefore, transmission from the urethra is largely dependent on the proportion of asymptomatic urethral gonorrhoea cases and also how quickly symptomatic individuals access health care. Indeed the majority (57%) of cases transmitted from the urethra are transmitted from the 20% of asymptomatic urethral infections. If the proportion of asymptomatic cases is higher or access to treatment is more difficult, then the urethra may become substantially more important for driving the population prevalence.27
Our findings that over 95% of rectal infections are transmitted from the oropharynx and only 5% from the urethra may at first seem surprising (Fig. 2). This finding is however sensitive to the proportion of asymptomatic urethral gonorrhoea or untreated cases. However, identifying exactly which sexual practices are responsible for transmission of NG from the oropharynx to the rectum is complex. Given that NG can be cultured in saliva, transmission from the oropharynx could occur through oral sex and then anal sex (saliva on penis), oroanal sex, or digital or anal penetration with saliva used as a lubricant even if condoms are used. Sexual surveys are yet to collect such detailed information and to date, none have been able to assess these acts separately.8 Importantly, however, observational studies have shown that condoms reduce the risk of rectal gonorrhoea but it is conceivable that this occurs through covering the saliva contaminated penis. This and the use of saliva as a lubricant for anal sex may explain why condoms are not 100% protective. It is difficult to explore this finding further until large studies with adequate sample size to stratify by risks including saliva use, riming, anal sex and condom use are conducted.
Our modeling suggests that most urethral infections (>65%) are acquired from the rectum rather than from the oropharynx, which is consistent with the observation that 30% of urethral NG is attributable to oral sex.10 We have reported relatively low per-act transmission from the rectum or oropharynx to the urethra. This finding contrasts with the relatively high per-act transmission probability from the cervix to the penis of about 25%.21 One possible explanation for this could be the much lower NG bacterial load at the oropharynx or anus compared to the cervix. Supporting this argument is the high sensitivity of cervical swabs for culture compared with NAAT but very low sensitivity of throat or anal samples.28
Our model supports the observation that despite increases in screening in MSM in Australia gonorrhoea notifications have not shown any signs of subsiding.2,17 It is therefore concerning that the model suggests that further increasing the frequency of screening will not dramatically change incidence.29 Further increases in screening may also be difficult to achieve.
Our model suggests that widespread use of mouthwash may substantially reduce gonorrhoea prevalence in MSM. There is good data supporting the use of antibacterial/antiseptic mouthwash products (eg, Listerine) to reduce a broad spectrum of oral bacteria including NG.6,24 Mouthwash is also commonly used by MSM in Australia, but few (28%) use it daily indicating a substantial room to increase use as an intervention (unpublished MSHC data). Our modelling suggests a 50% coverage of use of a “moderately effective” (1% reduction in infection duration per use) mouthwash among MSM can result in a 7-fold reduction in the prevalence of gonorrhoea and a “highly effective” (1.5% reduction in infection duration per use) mouthwash may achieve a scenario close to elimination (Figs. 3–4). The key to its effectiveness however will be identifying those brands that have activity against NG and ensuring these are used commonly.
We note a number of limitations with this study. First, this model assumed that behavioral changes and interventions had immediate effects on gonorrhoea prevalence. In reality, impacts of any interventions would require sometime to be reflected in the population. Second, coinfection of gonorrhoea at the different sites in the same person was not considered in the model, leading to a potential underestimate of transmission. Third, although we investigated transmission of NG with a mixing sexual behavioral model, variations in the other sexual behavioral patterns, such as the duration of sexual practices or group sex among MSM, may significantly contribute to the model uncertainties. Fourth, the study findings are limited by the current understanding of the pathophysiology of gonorrhoea; uncertainties in many indicators, such as the natural duration of oropharyngeal infection, the proportion of asymptomatic urethral infection, bacterial load at various anatomical sites, may affect the estimate of transmission probabilities. Fifth, the prevalence of gonorrhoea infection was estimated based on sexual health clinic attendees, which may be higher than the actual prevalence in community MSM population. Sixth, only coverage but not frequency of screening was accounted for in our current model. Seventh, we have assumed a linear effect on the mouthwash efficacy; in reality, the clearance due to mouthwash may be greater at the early infection than later.
Perhaps, the most significant limitation of our model is the paucity of data on sexual practices. We felt that despite this limitation, it was important to explore the complexities of gonorrhoea transmission between different anatomic sites among MSM because of the rapidly rising rates. On the basis of the available data and the findings of our model, we urge researchers to prove or disapprove the findings of the model through empirical epidemiological studies; before further extensive antibiotic resistance develops in NG.
1. Mulhall BP, Wright S, Allen D, et al. High Rates of Sexually transmitted infections in HIV-positive patients in the Australian HIV Observational Database—A prospective cohort study. Sex Health 2014; 11:291–297.
2. The Kirby Institute, HIV, viral hepatitis and sexually transmissible infections in Australia Annual Surveillance Report 2014. 2014, The Kirby Institute, The University of New South Wales: Sydney.
3. Golub SA, Kowalczyk W, Weinberger CL, et al. Preexposure prophylaxis and predicted condom use among high-risk men who have sex with men. J Acquir Immune Defic Syndr 2010; 54:548–555.
4. Hoff CC, Chakravarty D, Bircher AE, et al. Attitudes towards PrEP and anticipated condom use among concordant HIV-negative and HIV-discordant male couples. AIDS Patient Care STDS 2015; 29:408–417.
5. Lewis DA, Sriruttan C, Müller EE, et al. Phenotypic and genetic characterization of the first two cases of extended-spectrum-cephalosporin-resistant Neisseria gonorrhoeae
infection in South Africa and association with cefixime treatment failure. J Antimicrob Chemother 2013; 68:1267–1270.
6. Fairley CK, Hocking JS, Zhang L, et al. Frequent transmission of gonorrhea in men who have sex with men. Emerg Infect Dis 2017; 23:102–104.
7. Morris SR, Klausner JD, Buchbinder SP, et al. Prevalence and incidence of pharyngeal gonorrhea in a longitudinal sample of men who have sex with men: The EXPLORE study. Clin Infect Dis 2006; 43:1284–1289.
8. Jin F, Prestage GP, Mao L, et al. Incidence and risk factors for urethral and anal gonorrhoea and chlamydia in a cohort of HIV-negative homosexual men: The health in men study. Sex Transm Infect 2007; 83:113–119.
9. Chow EP, Camilleri S, Ward C, et al. Duration of gonorrhoea and chlamydia infection at the pharynx and rectum among men who have sex with men: A systematic review. Sex Health 2016; 13:199–204.
10. Barbee LA, Khosropour CM, Dombrowski JC, et al. An estimate of the proportion of symptomatic gonococcal, chlamydial and non-gonococcal non-chlamydial urethritis attributable to oral sex among men who have sex with men: A case–control study. Sex Transm Infect 2015.
11. Chow EP, Lee D, Tabrizi SN, et al. Detection of Neisseria gonorrhoeae
in the pharynx and saliva: implications for gonorrhoea transmission. Sex Transm Infect. 2015.
12. Chow EP, Cornelisse VJ, Read TR, et al. Saliva use as a lubricant for anal sex is a risk factor for rectal gonorrhoea among men who have sex with men, a new public health message: A cross-sectional survey. Sex Transm Infect. 2016.
13. Butler LM, Osmond DH, Jones AG, et al. Use of saliva as a lubricant in anal sexual practices among homosexual men. J Acquir Immune Defic Syndr, 2009; 50:162–167.
14. Chow EP, Howden BP, Walker S, et al. Antiseptic mouthwash against pharyngeal Neisseria gonorrhoeae
: A randomised controlled trial and an in vitro study. Sex Transm Infect 2017; 93:88–93.
15. Hui B, Fairley CK, Chen M, et al. Oral and anal sex are key to sustaining gonorrhoea at endemic levels in MSM populations: A mathematical model. Sex Transm Infect 2015; 91:365–369.
16. STI's in Gay Men Action Group (STIGMA). Sexual transmitted infection testing guidelines for men who have sex with men. STI's in Gay Men Action Group (STIGMA) Web site. 2014. 2014 [cited 2015 31 Dec]; Available from: http://stipu.nsw.gov.au/wp-content/uploads/STIGMA_Testing_Guidelines_Final_v5.pdf
17. Chow EP, Tomnay J, Fehler G, et al. Substantial increases in chlamydia and gonorrhea positivity unexplained by changes in individual-level sexual behaviors among men who have sex with men in an Australian sexual health service from 2007 to 2013. Sex Transm Dis 2015; 42:81–87.
18. Fairley CK, Chow EP, Hocking JS. Early presentation of symptomatic individuals is critical in controlling sexually transmissible infections. Sex Health 2015; 12:181–182.
19. Ryder N, Lockart IG, Bourne C. Is screening asymptomatic men who have sex with men for urethral gonorrhoea worthwhile? Sex Health 2010; 7:90–91.
20. Russell CJ, Golub SA, Cohen DE, et al. Urine-based asymptomatic urethral gonorrhea and chlamydia screening and sexual risk-taking behavior in men who have sex with men in greater Boston. AIDS Patient Care STDS 2007; 21:205–211.
21. Chen MI, Ghani AC. Populations and partnerships: insights from metapopulation and pair models into the epidemiology of gonorrhoea and other sexually transmitted infections. Sex Transm Infect 2010; 86(Suppl 3):63–69.
22. Hull P, Mao L, Kolstee J. Gay Community Periodic Survey: Sydney 2015. UNSW Australia: Sydney: Centre for Social Research in Health, 2015.
23. de Wit J, Mao L, Adam P, et al. HIV/AIDS, hepatitis and sexually transmissible infections in Australia: Annual report of trends in behaviour 2015. 2015, Centre for Social Research in Health. UNSW Australia: Sydney.
24. Netuschil L, Weiger R, Preisler R, et al. Plaque bacteria counts and vitality during chlorhexidine, meridol and listerine mouthrinses. Eur J Oral Sci 1995; 103:355–361.
25. Juher D, Ripoll J, Saldana J. Analysis and Monte Carlo simulations of a model for the spread of infectious diseases in heterogeneous metapopulations. Phys Rev E Stat Nonlin Soft Matter Phys 2009; 80(4 Pt 1):041920.
26. Rosenberger JG, Reece M, Schick V, et al. Sexual behaviors and situational characteristics of most recent male-partnered sexual event among gay and bisexually identified men in the United States. J Sex Med 2011; 8:3040–3050.
27. White PJ, Ward H, Cassell JA, et al. Vicious and virtuous circles in the dynamics of infectious disease and the provision of health care: Gonorrhea in Britain as an example. J Infect Dis 2005; 192:824–836.
28. Wind CM, de Vries HJ, Schim van der Loeff MF, et al. Successful combination of nucleic acid amplification test diagnostics and targeted deferred Neisseria gonorrhoeae
culture. J Clin Microbiol 2015; 53:1884–1890.
29. Guy R, Wand H, Franklin N, et al. Re-testing for chlamydia at sexual health services in Australia, 2004–08. Sex Health 2011; 8:242–247.