Previous studies have highlighted the importance of genital ulcer disease (GUD) in both the acquisition and transmission of the human immunodeficiency virus (HIV).1–5 Sexually transmitted infections (STIs) account for most of the GUD cases, although other infectious and noninfectious etiologies exist.6–8 Urethral pathogens, which may be cotransmitted with GUD pathogens or acquired independently, may cause either symptomatic or asymptomatic disease in patients with GUD.9 Clinical urethritis, particularly when inflammation is intense, may be associated with increased risk of HIV transmission.10,11
Syndromic management aims to ensure that common etiologic agents responsible for a given STI syndrome are treated appropriately. South Africa adopted the syndromic management approach in the late 1990s when chancroid was the leading cause of GUD.7,9 Algorithms were designed to treat GUD with a combination of intramuscular (i.m.) benzathine penicillin and oral erythromycin.12 During the past decade, South Africa has witnessed a dramatic increase in the prevalence of both HIV and genital herpes and the near elimination of chancroid.13 Accordingly, oral acyclovir was recently added to the existing first-line GUD therapy in revised national STI guidelines.14 For those men with both GUD and male urethritis syndrome (MUS), treatment can easily be modified to ensure satisfactory treatment of all common GUD and urethral discharge pathogens.14 However, in countries practicing STI syndromic management, surveillance, and/or research data are required to decide on the most cost-effective and clinically beneficial combination of antimicrobial agents, which will ensure that patients with GUD are treated for common asymptomatic STI pathogens as well as their presenting syndrome.
The primary objective of this study was to report the prevalence of GUD and urethral pathogen confections in South African men who participated in a randomized controlled trial (RCT) to assess the efficacy of episodic acyclovir as an additional first-line treatment of GUD.15 Secondary objectives were to investigate the associations of GUD and urethral pathogens with both HIV coinfection and clinical features.
MATERIALS AND METHODS
Men with GUD were recruited to an RCT to evaluate the effect of acyclovir episodic therapy on the ulcer duration and HIV lesional shedding. The trial was registered at ClinicalTrials.gov with identifier NCT00164424 and has been reported previously.15 The study was conducted at 3 public sector primary health care clinics in Gauteng Province, South Africa, from March 2005 to December 2006. Briefly, consenting men with GUD were recruited if they were 18 to 60 years and were willing to be tested for HIV antibodies and to return for follow-up visits. Participants with extensive ulceration (surface area > 500 mm2) or chronic ulcers (duration > 1 month) were excluded. A baseline questionnaire was administered to record demographic, clinical, and behavioral data and biologic specimens (blood, ulcer swabs, first-pass urine) were collected. Study nurses were trained in all study procedures, including genital examination, visualization of urethral discharge (including urethral squeezing in cases where discharge was not evident on initial inspection), recording of ulcer characteristics, and measurement of ulcer surface area.
Sexually Transmitted Infection Treatment for Participants
Participants received free STI treatment, partner notification slips, condoms, and health education. As per protocol, participants received single-dose antibiotic therapy for GUD with i.m. benzathine penicillin 2.4 MU and oral ciprofloxacin 500 mg (given in lieu of erythromycin). Participants randomly received either acyclovir (400 mg, once in every 8 hours for 5 days) or placebo with a similar dosage schedule. Participants with MUS also received oral doxycycline (100 mg, once in every 12 hours for 7 days) at baseline. As single-dose ciprofloxacin 500 mg was national first-line therapy for gonorrhea in 2005 to 2006, positive Neisseria gonorrhoeae test results were only re-treated with a single-dose i.m. ceftriaxone 250 mg if quinolone resistance was suspected on clinical grounds. Asymptomatic chlamydial and Mycoplasma genitalium infections were also treated with oral doxycycline (as per MUS cases), whereas trichomoniasis was treated with oral metronidazole (400 mg, once in every 12 hours for 7 days).
Specimen Collection for STI/HIV Diagnostic Testing
At the initial visit, a specimen to determine ulcer etiology was obtained by rotating a Dacron swab in the ulcer base. Ulcer swabs were stored at −70°C until tested. First-pass urine (20–30 mL) was collected to screen participants for urethral pathogens. Blood was obtained for rapid HIV testing (finger prick) as well as for serologic testing to detect syphilis and HSV-2 infection. Unless otherwise stated, laboratory tests were performed at the Centre for HIV and STIs, National Institute for Communicable Diseases, Johannesburg, South Africa.
Detection of Ulcer Pathogens
A real-time multiplex polymerase chain reaction (M-PCR) assay on a Rotor-Gene 3000 platform (Corbett Robotics Pty Ltd., Sydney, Australia) was used to detect HSV, Treponema pallidum and Haemophilus ducreyi in DNA extracted from the ulcer swabs using the Xtractor Gene (Corbett Robotics Pty Ltd.). The primers and probes targeted the genes encoding the HSV glycoprotein D protein, the T. pallidum 47-kd lipoprotein and the H. ducreyi HhdA haemolysin protein. After incubation at 50°C (2 minutes) and Taq activation at 95°C (10 minutes), 40 cycles of denaturation (95°C, 20 seconds) and annealing/extension (60°C, 60 seconds) took place. For HSV DNA–positive cases, a specific in-house duplex real-time PCR assay using a Rotor-Gene 3000 platform (Corbett Robotics Pty Ltd.), targeting the HSV glycoprotein G, was used to type HSV DNA as either HSV-1 or HSV-2 at the Centers for Disease Control and Prevention (CDC, Atlanta, GA).13 An in-house real-time PCR, developed by Morré et al.,16 was used to detect Chlamydia trachomatis L1-L3, the causative agent of lymphogranuloma venereum (LGV). The primers and probe targeted the pmpH gene encoding the polymorphic membrane protein H of C. trachomatis L2b. After incubation at 50°C (2 minutes) and Taq activation at 95°C (10 minutes), 40 cycles of denaturation (95°C, 15 seconds) and annealing/extension (60°C, 60 seconds) took place. Extracted genomic DNAs from the following strains were used as controls: HSV-2 (ATCC VR-540), T. pallidum Nichol’s strain (purified genomic DNA from rabbit liver, courtesy of Dr. Mike Norgard, Dallas, TX), H. ducreyi (ATCC 700724), C. trachomatis strain 440 (ATCC VR-901B, L1 serovar), C. trachomatis strain T’ang (TRIC/China/Peking-2/OTf, ATCC VR-577, L2 serovar) and C. trachomatis strain 404 (ATCC VR-903, L3 serovar).
Detection of Urethral Pathogens
A real-time M-PCR assay, also using a Rotor-Gene 3000 platform, was used to detect N. gonorrhoeae, C. trachomatis, Trichomonas vaginalis and M. genitalium in DNA extracted from first-pass urine using the Xtractor Gene (Corbett Robotics Pty Ltd.). The primers and probes targeted the N. gonorrhoeae cytosine-specific DNA methyltransferase gene, the cryptic plasmid of C. trachomatis, the T. vaginalis repeated DNA fragment, and the M. genitalium pdhD gene, encoding for dihydrolipoamide dehydrogenase. After incubation at 50°C (2 minutes) and Taq activation at 95°C (10 minutes), 40 cycles of denaturation (95°C, 20 seconds) and annealing/extension (60°C, 60 seconds) took place. Genomic DNA extracts prepared from the following ATCC strains served as controls: N. gonorrhoeae (ATCC 700825), C. trachomatis (ATCC VR-885), T. vaginalis (ATCC 30001), and M. genitalium (ATCC 33530).
Human Immunodeficiency Virus Rapid and Serologic Testing
Human immunodeficiency virus testing was done on site with 2 rapid HIV tests (Determine; Inverness Medical Japan Co., Ltd., Tokyo, Japan; Capillus; Trinity Biotech Plc, Bray, Ireland), supported by pretest and posttest counseling. Sera were used to test for HSV-2 using a type-specific serologic test (Kalon Biological Ltd., Guildford, United Kingdom) and for syphilis using a rapid plasma reagin (RPR) assay (Macro-Vue RPR, Becton Dickinson and Co.) and a T. pallidum particle agglutination assay (TPPA) (Serodia-TPPA, Fujirebio Inc., Japan).
We conducted exploratory analyses to investigate variables associated with the detection of genital ulcer and urethral pathogens. Significance of associations was determined using the χ2 test. All statistical analyses were performed using Statistical Analysis Software (SAS) version 9.1 (SAS Institute Inc., Cary, NC). With reference to a criterion standard (detection of urethral pathogens by real-time M-PCR), the sensitivity, specificity, positive predictive value, and negative predictive value of 2 clinical variables (dysuria, nurse-observed urethral discharge) for predicting infection with “any urethral pathogen” were calculated using standard formulae.
Ethics and Other Approvals
The study was approved by the Ethics Committees of the University of the Witwatersrand (South Africa), the London School of Hygiene and Tropical Medicine (United Kingdom) and the Institutional Review Board of the CDC (United States). Approvals were also obtained from the Gauteng Provincial Department of Health and the City of Johannesburg Health Department.
As described previously, 1559 men were screened, 635 were deemed eligible to participate, and 615 were recruited in the RCT and thus included in this analysis.15 The main reasons for ineligibility were presentation with healed ulcers, unwillingness to have an HIV test, and inability to attend follow-up appointments.15 The median age of participants was 29 years (interquartile range Q1–Q3, 25–35 years). A history of GUD or urethral discharge was reported by 252 (41.0%) and 151 (24.6%) men, respectively.
Pathogens Isolated From Genital Ulcers
Laboratory data were available for 613 men with GUD (99.7%). As reported previously, genital herpes was the pathogen most frequently detected in the ulcer swabs (451 men, 73.6%).15 Herpes simplex virus 2 was present in 443 (98.2%) of these herpetic ulcers; HSV-1 was present in 19 (4.2%) of herpetic ulcers, either as a single pathogen (5 cases, 1.1%) or in combination with HSV-2 (14 cases, 3.1%). It was not possible to perform HSV typing on 3 cases (0.7%) of proven genital herpes. Bacterial pathogens were detected in 48 ulcers (7.8%); 12 of these ulcers (10 syphilis cases, 2 LGV cases) were coinfected with either HSV-1 (3 cases), HSV-2 (4 cases), or both HSV types (5 cases). Syphilis was the most frequently detected bacterial ulcerative disease (30 cases, 4.9%), followed by chancroid (10 cases, 1.6%) and LGV (8 cases, 1.3%). No ulcers had more than 1 bacterial agent identified. No pathogens were detected in 126 ulcers (20.6%).
Prevalence of Urethral Pathogens
Among the 615 participants, 188 (30.6%) had at least 1 urethral pathogen detected and 30 (4.9%) had multiple pathogens. In all, there were 223 urethral infections diagnosed by M-PCR, including 69 (11.2%) M. genitalium, 64 (10.4%) T. vaginalis, 60 (9.8%) C. trachomatis, and 30 (4.9%) N. gonorrhoeae infections.
Anti-HIV and anti–HSV-2 antibodies were detected in 387 (62.9%) and 434 (70.6%) of 615 men, respectively. Syphilis serologic results were available for 613 men (99.7%) overall and for 611 men (99.3%) for whom GUD pathogen data were available. Overall, few men (54, 8.8%) had a positive RPR test result. RPR seropositivity was significantly associated with both the detection of T. pallidum and the failure to detect HSV in the genital ulcer (P < 0.001 for both). Most positive RPR test results (46/54, 85.2%) were confirmed by a positive TPPA assay result. Overall, 141 men (23.0%) were TPPA seropositive, indicating lifetime exposure to T. pallidum. All 30 patients with ulcers positive for T. pallidum by M-PCR were seropositive with the TPPA assay and 20 (69.0%) of 29 patients with sera tested from these primary syphilis cases were RPR seroreactive.
Associations Between Genital Ulcer Pathogen Isolation and Clinical Features
Self-reported pain and ulcer tenderness on examination were associated with bacterial infections (including mixed infections) rather than herpetic ulceration alone (P = 0.027 and P = 0.041, respectively) (Table 1). When analyzed by individual bacterial pathogen isolated (data not shown), tenderness was only associated with T. pallidum infection (P = 0.010). Herpetic ulcers were more likely to be superficial (P = 0.009), whereas syphilis and chancroid ulcers were more likely to be deep (P = 0.004 and P < 0.001, respectively). The presence of an irregular ragged margin (P = 0.042) and a purulent ulcer base (P = 0.014) were also associated with chancroid.
Associations Between Urethral Pathogen Detection and MUS Clinical Features
Dysuria was reported by 170 (27.6%) of 615 participants, whereas nurses observed urethral discharge in 69 (11.5%) of 601 participants (Table 2). Dysuria and nurse-observed urethral discharge was associated with N. gonorrhoeae (P < 0.001 for both) and T. vaginalis infections (P = 0.014 and P < 0.001, respectively) but not with the presence of other urethral pathogens. Urethral pathogens were detected in 102 (24.9%) of 409 men without symptoms of dysuria or nurse-confirmed urethral discharge (Table 2). This proportion represents 56.7% of all men in whom urethral pathogens were detected. With a criterion standard of M-PCR detection of “any urethral pathogen,” dysuria and nurse-observed urethral discharge each had a low sensitivity (36.7% and 21.9%, respectively), specificity (76.3% and 93.1%, respectively), positive predictive value (40.6% and 58.0%, respectively), and negative predictive value (73.3% and 73.1%, respectively) for predicting patients’ infection with urethral pathogens.
Associations Between GUD/Urethral Pathogens and HIV Serostatus
Almost two thirds (387, 62.9%) of participants were HIV seropositive at baseline. The detection of T. pallidum in ulcers was associated with a negative HIV serostatus (18 cases per 209 HIV seronegative participants vs. 12 cases per 387 HIV seropositive participants, P = 0.007). There were no associations between HIV serostatus and the detection of other GUD pathogens. Among urethral pathogens, HIV seropositivity was only associated with the presence of M. genitalium (P = 0.023).
This study, the largest sub–Saharan Africa–based genital ulcer study in recent years, reports several clinically relevant findings. First, our study has confirmed the predominance of HSV, in particular HSV-2, and the comparatively low prevalence of bacterial pathogens among men with GUD. Second, both clinical features of MUS and asymptomatic urethral pathogens were frequently detected. Finally, few GUD or urethral pathogens were associated with HIV serostatus.
Early diagnosis and effective treatment of GUD and urethral discharges may reduce morbidity and decrease HIV transmission.1,4,17 To achieve this, syndromic management algorithms should include antimicrobials that cover common STI pathogens and be informed by annual laboratory-based surveillance data. Surveillance initiatives remain weak in Africa, resulting in continued emphasis on outdated treatment algorithms to manage curable bacterial ulcerative conditions that were highly prevalent a decade ago.7,9,18 The World Health Organization has, for a number of years, recommended the inclusion of antiherpes medication in the GUD treatment algorithms of countries where genital herpes is responsible for 30% or more of GUD cases.19 However, few countries have taken up this recommendation despite the existence of cost-effectiveness modeling data.20 Our observation that genital herpes accounts for almost three quarters of all cases of GUD supports other African studies.21 The dramatic decrease in the relative prevalence of chancroid may be attributed in part to the syndromic management approach.22,23
In view of the substantial decline in the relative prevalence of bacterial GUD cases, it would be advantageous to limit unnecessary injections of benzathine penicillin and overdispensing of erythromycin. Hoyo et al.24 have suggested that unweighted algorithms using basic clinical information may be used to improve diagnostic accuracy of GUD. However, our study found that clinical features of ulcers were generally poor indicators of GUD etiology. The commercial development of affordable, sensitive, and specific rapid diagnostic tests for a range of ulcer pathogens may aid rational treatment of GUD in the future.25–29
At present, there is a lack of guidance to support removal of antibiotics for low-prevalence ulcer pathogens from treatment guidelines, in part owing to concern that these pathogens may reemerge undetected in countries where laboratory-based STI surveillance is suboptimal and sporadic. Even in countries with functional surveillance systems, the question as to what pathogen prevalence and how many studies should guide removal of specific antibiotics remains unanswered. Similarly, there is no consensus as to what coinfecting pathogen prevalence and how many confirmatory studies should trigger inclusion of additional antibiotics into the GUD treatment package.
This is the third study to report high prevalence of coexistent urethral pathogens among South African men. The first study, undertaken in 1998 among 186 miners with genital ulcers but without urethritis symptoms, demonstrated that 53% had urethritis based on microscopy, 24% had gonorrhea, 6% had chlamydial infection, and 7% had M. genitalium infection.9 The second study took place at an STI clinic in Johannesburg between 2004 and 2006 in which 135 male patients with GUD of undetermined etiology were enrolled.30 Among the 76 asymptomatic men (56%), 16% had gonorrhea, 13% had chlamydial infection, 13% had M. genitalium, and 20% had trichomoniasis. The higher prevalence of gonorrhea reported among the miners likely reflects their higher relative risk for STI acquisition, as supported by the observation that chancroid accounted for more than 50% of their ulcers.9
Patients with GUD presenting with clinical features of MUS should receive appropriate syndromic treatment. However, asymptomatic urethral pathogens may remain inadequately treated by GUD-specific therapy. Although the 1 week’s course of erythromycin given for chancroid and LGV in South Africa’s GUD flowchart will cover both chlamydial and, to some extent, M. genitalium infection, it will not treat T. vaginalis infection and it is not ideal for the treatment of gonorrhea, owing to concerns over the emergence of macrolide resistance. Our prevalence data for urethral pathogen coinfections provide support for the continued use of erythromycin in the GUD algorithm and, once more, raise the question as to whether patients with GUD should receive additional antibiotics to cover asymptomatic T. vaginalis and N. gonorrhoeae infections.9,30
The observation that most ulcer and urethral pathogens were not associated with HIV infection is of interest, given that STIs are frequently reported as risk factors for HIV transmission.1–4 Rottingen et al.5 have already pointed out the publication bias and gaps in the focus of studies assessing HIV-STI interactions in the context of heterosexual HIV epidemics and argue for a better understanding of these interactions.
Our study had some limitations. First, as an RCT, it selected for men who were willing to have an HIV test and attend several follow-up visits. Second, men alone were recruited, and our findings may not be relevant for women. Third, our study took place within Gauteng Province and may not be nationally representative. Lastly, the prevalence of donovanosis was not assessed, although this may not be important because other contemporary GUD etiologic surveys in Johannesburg did not detect any cases (D. Lewis, NICD/NHLS, Johannesburg, unpublished data).
In conclusion, our study has highlighted the importance of genital herpes, and these data supported the National Department of Health’s decision to include acyclovir as part of first-line GUD therapy in the 2008 revision of South Africa’s STI treatment guidelines.14 In addition, our study has again demonstrated a significant burden of urethral pathogens in patients with GUD, and these data require further consideration.
The prevalence data on GUD pathogens were previously reported by HIV serostatus in the main trial article.15
1. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: The contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75: 3–17.
2. Gray RH, Wawer MJ, Sewankambo NK, et al.. Relative risks and population attributable fraction of incident HIV associated with symptoms of sexually transmitted diseases and treatable symptomatic sexually transmitted diseases in Rakai District, Uganda. Rakai Project Team. AIDS 1999; 13: 2113–2123.
3. Plummer FA, Simonsen JN, Cameron DW, et al.. Cofactors in male-female sexual transmission of human immunodeficiency virus type 1. J Infect Dis 1991; 163: 233–239.
4. Dickerson MC, Johnston J, Delea TE, et al.. The causal role for genital ulcer disease as a risk factor for transmission of human immunodeficiency virus. An application of the Bradford Hill criteria. Sex Transm Dis 1996; 23: 429–440.
5. Rottingen JA, Cameron DW, Garnett GP. A systematic review of the epidemiologic interactions between classic sexually transmitted diseases and HIV: How much really is known? Sex Transm Dis 2001; 28: 579–597.
6. Behets FM, Brathwaite AR, Hylton-Kong T, et al.. Genital ulcers: Etiology, clinical diagnosis, and associated human immunodeficiency virus infection in Kingston, Jamaica. Clin Infect Dis 1999; 28: 1086–1090.
7. Morse SA, Trees DL, Htun Y, et al.. Comparison of clinical diagnosis and standard laboratory and molecular methods for the diagnosis of genital ulcer disease in Lesotho: Association with human immunodeficiency virus infection. J Infect Dis 1997; 175: 583–589.
8. Risbud A, Chan-Tack K, Gadkari D, et al.. The etiology of genital ulcer disease by multiplex polymerase chain reaction and relationship to HIV infection among patients attending sexually transmitted disease clinics in Pune, India. Sex Transm Dis 1999; 26: 55–62.
9. Ballard RC, Fehler HG, Htun Y, et al.. Coexistence of urethritis with genital ulcer disease in South Africa: Influence on provision of syndromic management. Sex Transm Infect 2002; 78: 274–277.
10. Winter AJ, Taylor S, Workman J, et al.. Asymptomatic urethritis and detection of HIV-1 RNA in seminal plasma. Sex Transm Infect 1999; 75: 261–263.
11. Cohen MS, Hoffman IF, Royce RA, et al.. Reduction of concentration of HIV-1 in semen after treatment of urethritis: Implications for prevention of sexual transmission of HIV-1. AIDSCAP Malawi Research Group. Lancet 1997; 349: 1868–1873.
12. National Department of Health. Essential drug list and standard treatment guidelines —primary health care (PHC). 2nd ed. Pretoria, South Africa: National Department of Health, 2003.
13. Lai W, Chen CY, Morse SA, et al.. Increasing relative prevalence of HSV-2 infection among men with genital ulcers from a mining community in South Africa. Sex Transm Infect 2003; 79: 202–207.
14. National Department of Health. First line comprehensive management and control of sexually transmitted infections (STIs): Protocol for the management of a person with a sexually transmitted infection according to the Essential Drug List. Pretoria, South Africa: National Department of Health, 2008.
15. Paz-Bailey G, Sternberg M, Puren AJ, et al.. Improvement in healing and reduction in HIV shedding with episodic acyclovir therapy as part of syndromic management among men: A randomized, controlled trial. J Infect Dis 2009; 200: 1039–1049.
16. Morre SA, Spaargaren J, Fennema JS, et al.. Real-time polymerase chain reaction to diagnose lymphogranuloma venereum. Emerg Infect Dis 2005; 11: 1311–1312.
17. Grosskurth H, Mosha F, Todd J, et al.. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: Randomised controlled trial. Lancet 1995; 346: 530–536.
18. Behets FM, Liomba G, Lule G, et al.. Sexually transmitted diseases and human immunodeficiency virus control in Malawi: A field study of genital ulcer disease. J Infect Dis 1995; 171: 451–455.
20. Vickerman P, Ndowa F, Mayaud P. Modelling the cost per ulcer treated of incorporating episodic treatment for HSV-2 into the syndromic algorithm for genital ulcer disease. Sex Transm Infect 2008; 84: 243–248.
21. Paz-Bailey G, Rahman M, Chen C, et al.. Changes in the etiology of sexually transmitted diseases in Botswana between 1993 and 2002: Implications for the clinical management of genital ulcer disease. Clin Infect Dis 2005; 41: 1304–1312.
22. Htun Y, Morse SA, Dangor Y, et al.. Comparison of clinically directed, disease specific, and syndromic protocols for the management of genital ulcer disease in Lesotho. Sex Transm Infect 1998; 74 (suppl 1): S23–S28.
23. Cheluget B, Joesoef MR, Marum LH, et al.. Changing patterns in sexually transmitted disease syndromes in Kenya after the introduction of a syndromic management program. Sex Transm Dis 2004; 31: 522–525.
24. Hoyo C, Hoffman I, Moser BK, et al.. Improving the accuracy of syndromic diagnosis of genital ulcer disease in Malawi. Sex Transm Dis 2005; 32: 231–237.
25. Dangor Y, Ballard RC, da L Exposto F, et al.. Accuracy of clinical diagnosis of genital ulcer disease. Sex Transm Dis 1990; 17: 184–189.
26. Fast MV, D’Costa LJ, Nsanze H, et al.. The clinical diagnosis of genital ulcer disease in men in the tropics. Sex Transm Dis 1984; 11: 72–76.
27. O’Farrell N, Hoosen AA, Coetzee KD, et al.. Genital ulcer disease: Accuracy of clinical diagnosis and strategies to improve control in Durban, South Africa. Genitourin Med 1994; 70: 7–11.
28. Peeling RW, Holmes KK, Mabey D, et al.. Rapid tests for sexually transmitted infections (STIs): The way forward. Sex Transm Infect 2006; 82 (suppl 5): v1–v6.
29. Pettifor A, Walsh J, Wilkins V, et al.. How effective is syndromic management of STDs?: A review of current studies. Sex Transm Dis 2000; 27: 371–385.
30. Black V, Magooa P, Radebe F, et al.. The detection of urethritis pathogens among patients with the male urethritis syndrome, genital ulcer syndrome and HIV voluntary counselling and testing clients: Should South Africa’s syndromic management approach be revised? Sex Transm Infect 2008; 84: 254–258.