SEXUALLY TRANSMITTED DISEASES (STDs) have a major impact on morbidity and mortality in many countries, 1,2 partially because of their ability to enhance the rate of sexual transmission of HIV. 3–5 However, clinicians find the clinical diagnosis of genital ulcers difficult because the appearance of ulcers can mislead even expert clinicians, 6 and diagnostic tests are least available in settings in which the rates of disease are highest.
In response, the World Health Organization (WHO) has developed protocols for the syndromic management of genital ulcers in resource-poor settings. 7,8 However, because risk factors for sexually transmitted infections, 9 patterns of disease, 10 etiologies, 11 and antimicrobial susceptibilities 12–15 differ from region to region and change over time, assessment of such guidelines and adaptation to the local situation are necessary.
This study had three objectives: first, to determine the prevalence of Treponema pallidum, Haemophilus ducreyi, and herpes simplex virus (HSV) in genital ulcers with a duration of ≤60 days among male STD clinic attendees in Santo Domingo, Dominican Republic, and Lima, Peru; second, to model the performance of algorithms for the syndromic management of genital ulcers in each population; and third, to compare the detection of the three major genital ulcer pathogens by polymerase chain reaction (PCR) testing with other diagnostic alternatives for genital ulcer disease (GUD).
In Santo Domingo, consecutive male patients with genital ulcers of ≤60 days’ duration who presented at the Centro Dermatológico were enrolled between July 1995 and June 1996. In Lima, similar participants who attended the Centro Antivenéreo de Lima were evaluated between November 1994 and May 1995.
Interview and Examination
All patients with genital ulcers in Santa Domingo and Lima underwent the same interview in Spanish and the same examination, with use of standardized questionnaires and examination forms. Protocols were approved by the Human Subjects Research Committees of the University of Washington and the Investigational Review Board of the Centro Dermatovenereologia of Santo Domingo. Participants gave verbal informed consent; provided demographic information and medical and sexual histories; and described the nature of the current illness, including the duration of lesions and history of treatment. Physicians experienced in the diagnosis of STD recorded the characteristics of the lesions, the presence of adenopathy, and their clinical impressions of the probable cause of the lesions. Participants in both countries underwent multiplex PCR testing of genital lesions for T pallidum, H ducreyi, and HSV and serologic tests for HIV infection. Serologic tests for syphilis and other tests of ulcer specimens differed by country.
Specimen Collection and Laboratory Methods
After informed consent and pretest counseling, 15 ml of blood was obtained from all study participants for serologic tests. Tests for antibodies to HIV were performed by enzyme-linked immunosorbent assay (Abbott Laboratories, Abbott Park, IL) and confirmed by Western blot (Cambridge Biotech, Worcester, MA), according to the manufacturers’ specifications. Serologic diagnosis of HIV infection was based on WHO criteria. All sera from patients in the Dominican Republic were tested for syphilis by VDRL tests; reactive sera were further tested for VDRL-reactive titer, and findings were then confirmed by means of fluorescent treponemal antibody (absorbed) test (FTA-ABS; Difco Laboratories, Detroit, MI). In Peru, patients underwent quantitative rapid plasma reagin (RPR) tests (Organon Teknica, Turnhout, Belgium), with confirmatory testing by MHATP (Fujirebio, Japan).
Genital lesion specimen collection and multiplex PCR testing.
After thorough cleansing of each genital lesion with a sterile gauze pad, lesions were compressed to obtain exudate for dark-field examination. Material from the base of the largest genital lesion was collected on a Dacron swab for PCR testing. The swabs were placed into Digene specimen transport medium, contained in the Digene Specimen Collection Kit (Digene, Beltsville, MD), and frozen at −70 °C until analysis at the University of Washington for T pallidum, H ducreyi, and HSV DNA with the Roche Molecular Systems Multiplex PCR assay (M-PCR; Roche Molecular Systems, Pleasanton, CA), as previously described, 16 except that 10 U rather than 2 U of Taq polymerase (AmpliTaq; Perkin-Elmer, Norwalk, CT) was used per PCR reaction. In addition, samples were diluted 10-fold in AMPLICOR specimen transport medium (Roche Molecular Systems) before analysis to accommodate the collection medium (Digene STM), which differed from the Roche AMPLICOR specimen transport medium for which the test was developed. As described for the M-PCR procedure, all samples were tested in duplicate and analyzed for PCR inhibition with use of an internal control. Because of the higher proportion of PCR-negative specimens from Peru, all negative Peruvian genital ulcer specimens, and any inhibited specimens from the Dominican Republic were analyzed further after phenol extraction. No more positives were detected by this technique.
Additional testing for H ducreyi, T pallidum, and HSV.
In the Dominican Republic, each genital ulcer sample was inoculated onto a V agar plate, which contains 10% fetal bovine serum and 3 μg vancomycin, 17 and the plates were incubated at 33 °C to 35 °C in a candle jar for 2 to 3 days, checked for growth, and then incubated for 1 week before being discarded as negative. Possible H ducreyi isolates were subcultured to fresh charcoal medium plates, containing chocolate agar with 1% charcoal, 17 incubated 2 to 3 days, and then inoculated heavily with a cotton swab onto two agar media for X and V factor testing, as described. 18 Isolates with characteristic Gram stain appearance that grew around the X but not the V disk were presumptively identified as H ducreyi. H ducreyi isolates were confirmed in Seattle by their inability to synthesize porphyrins from 5-aminolevulinic acid using a porphyrin reagent kit (Remel, Lenexa, KS) per the manufacturer's directions, and by homology to whole cell DNA from the type strain of H ducreyi using the “taxonomic spot blot test.”18
In Lima, specimens were also tested for T pallidum by direct immunofluorescence with use of fluorescein-conjugated monoclonal antibiotics, 19 and for HSV with use of Herpchek (Dupont Laboratories, Billarica, MA). In the Dominican Republic, samples were examined by dark-field microscopy for detection of T pallidum.
Definition of “Gold Standards”
Diagnoses of syphilis, chancroid, and genital herpes were based on a positive M-PCR test for the causative agent. Serum microimmunofluorescence antibody titers determined at the University of Washington that were ≥1:512 for lymphogranuloma venereum (LGV) serovar 1 or 2 were considered indicative of acute LGV. Donovanosis was diagnosed presumptively by biopsy, with histologic sections stained with Giemsa; positive biopsy specimens were sent to Seattle for confirmation.
Modeling of Syndromic Management Algorithms for Genital Ulcer
The results of clinical examinations for genital ulceration and vesicular lesions and of M-PCR tests for T pallidum, H ducreyi, and HSV in genital lesions were used to retrospectively assess potential sensitivity, specificity, and positive predictive values (in terms of treatment of chancroid and syphilis) of a published WHO flow chart (algorithm) for syndromic management of genital ulcers. 7,8 A modified algorithm analogous to one previously published 20 (Figure 1) was then created and similarly assessed, in which only those patients with genital ulcers but no vesicular lesions were treated for chancroid and syphilis.
Sensitivity, specificity, and positive predictive values for different approaches to diagnosis of genital ulcers were estimated, and the Fisher exact test was used for statistical comparison.
Although the mean duration of symptoms was similar (approximately 2 weeks) in the two countries, patients in Peru were older than those in the Dominican Republic (mean, 40 years versus 31 years), had genital ulcers more frequently and used antibiotics less frequently, and had a somewhat different distribution of causative pathogens (Table 1). Therefore, findings are shown separately for the Dominican and Peruvian patients in all tables.
Results of M-PCR Testing
Eleven (14%) of 81 men in Santo Domingo and 2 (3%) of 63 in Lima were HIV-positive. In comparison, among Dominican women attending this clinic and family planning clinics, the concurrent HIV prevalence ranged from 1.9% for women with vaginal discharge to 5.5% among female sex workers; the prevalence of HIV among female sex workers attending STD clinics in Lima at the time of the study was 1.1% (J. Sanchez, unpublished data). Among the 81 patients in Santo Domingo, 4 (5%) were M-PCR-positive for T pallidum, 21 (26%) were positive for H ducreyi, and 35 (43%) were positive for HSV (Table 1). No pathogens were detected by PCR in 21 (26%). Of the 63 patients in Lima, 6 (10%) were M-PCR-positive for T pallidum, 24 (38%) were positive for HSV, and 3 (5%) were positive for both HSV and H ducreyi. No pathogens were detected in 30 cases (48%). Thus, significantly fewer patients had chancroid in Lima (5%) than in Santo Domingo (26%;P = 0.007).
The duration of symptoms of genital ulcers was not associated with M-PCR results for H ducreyi or T pallidum (data not shown) but was significantly associated with M-PCR results for HSV: the mean durations of symptoms for HSV M-PCR–positive and HSV M-PCR–negative persons were 10.5 and 15.7 days, respectively, in the Dominican Republic (P = 0.01) and 9.7 and 14.7 days, respectively, in Peru (P = 0.01).
History of antibiotic use did not reduce the detection of GUD pathogens by M-PCR. Thirty percent of Dominicans and 27% of Peruvians reported using antibiotics in the previous 48 hours, and 59% of Dominicans and 33% of Peruvians had used medications for the current episode since the onset of symptoms. History of antibiotic use for the current episode of GUD was not associated with decreased detection by M-PCR of T pallidum (M-PCR-positive: 5 of 73 with history of antibiotic use versus 5 of 70 without) or HSV (M-PCR-positive: 27 of 73 with history of antibiotic use versus 34 of 71 without), but such a history was associated with a trend toward increased detection of H ducreyi (detected in 17 of 73 with a history of antibiotic use and 7 of 70 without;P = 0.06).
Clinical Findings in Relation to M-PCR Results
At both sites the duration of symptoms before evaluation averaged approximately 2 weeks. The duration of symptoms was 40 days for the patient with biopsy-confirmed donovanosis—longer than the mean duration of other types of genital ulcers. History of ≥2 episodes of genital ulcers during the past year was associated with genital herpes. All 10 Dominicans who reported having ≥2 episodes of genital ulcers during the past year were M-PCR-positive for HSV, 9 of 16 Peruvians reporting ≥2 episodes of genital ulcers during the past year were M-PCR-positive for HSV (including 2 who were M-PCR-positive for H ducreyi as well), and another 2 were positive for HSV by EIA testing only. Nine of the 29 Peruvians with negative M-PCR tests for all 3 pathogens reported having ≥2 episodes of genital ulcers in the past year, suggesting that many of these patients also may have had herpes. As expected, the presence of vesicular lesions was significantly associated with detection of HSV by M-PCR: vesicular lesions were detected in 24 (39%) of 62 with HSV versus 8 (10%) of 82 without HSV by M-PCR (P < 0.01).
Comparison of PCR Results With Other Diagnostic Test Results
The M-PCR results, considered the “gold standard” in these analyses, are compared with results of other commonly used diagnostic methods in Table 2. All patients who were M-PCR-positive for T pallidum had reactive syphilis serologies and/or had a positive dark-field or direct immunofluorescence test for T pallidum. One M-PCR-negative patient (who had not taken antibiotics) had both a positive dark-field examination and positive serology (VDRL, 1:1; reactive FTA-ABS test).
Of 62 patients who were M-PCR-positive for HSV, 1 had reactive syphilis serology and 1 had a C trachomatis microimmunofluorescence serum antibody titer of 1:512, suggestive of LGV. The HSV EIA (Herpchek) used in Peru had a sensitivity of 70.4% (19/27), specificity of 85.3% (29/34), and positive predictive value of 79.2% (19/24) in comparison with M-PCR.
Presumptive Diagnosis Based on Clinical Findings
Clinicians in resource-poor settings often manage cases by making a presumptive diagnosis based on clinical findings or results of locally available laboratory tests or by treating syndromically according to an algorithm. For this study we contrasted these options against the laboratory gold standard (microbiological detection by M-PCR). Tables 3 and 4 compare presumptive clinical diagnoses with M-PCR results. The combined sensitivities of clinical diagnoses of either chancroid or syphilis were only 52% in the Dominican Republic and 22% in Lima; positive predictive values for either of these clinical diagnoses were only 32% in the Dominican Republic and 6% in Lima (Table 4). A clinical diagnosis of chancroid was common among patients with positive HSV tests by M-PCR. Of the 23 patients treated for syphilis on the basis of clinical diagnosis, 19 (83%) had a negative M-PCR test for syphilis; of the 49 treated for chancroid, 38 (78%) had a negative M-PCR for chancroid. Although the sensitivity and positive predictive value of clinical diagnosis of genital herpes were somewhat higher in both countries, even for this relatively common condition, performance of clinical diagnosis was poor. These findings are consistent with the conventional wisdom that genital ulcers are difficult to diagnose by sight, even for experienced clinicians.
Performance of a WHO Algorithm and a Modified Algorithm for Management of Genital Ulcer Disease
The theoretical performance of flow charts (algorithms) for management of GUD differed for these two patient populations with different prevalences of disease. Table 5 shows the sensitivity, specificity, and predictive values of the WHO algorithm for genital ulcers 8,9 in male patients with STD in Santo Domingo and Lima. In this algorithm, specificity for syphilis and chancroid is 0% and sensitivity for syphilis and chancroid is 100%, because with this algorithm all patients receive antibacterial treatment for both syphilis and chancroid. However, this algorithm had a positive predictive value of 31% for Dominican patients and only 14% for Peruvian patients. The percentage overtreated for syphilis with the WHO algorithm would be 95% in the Dominican Republic and 90% in Lima. Although rapid tests for syphilis (e.g., RPR) are available in clinical laboratories in Latin America, they are less widely used in primary care clinics for diagnostic testing for primary syphilis. In this study, of 10 patients whose M-PCR tests were positive for T pallidum, 8 (80%) had VDRL or RPR titers ≥1:4. If treatment for syphilis were reserved only for such seroreactive patients, sensitivity for syphilis would be 80%, specificity would be 97%, the positive predictive value would be 67%, and the rate of overtreatment for syphilis would be only 2.7%. The percentage overtreated for chancroid (number treated for syphilis who did not have syphilis, divided by total number studied) would be 74% in the Dominican Republic and 95% in Lima. Because chancroid was more common in the Dominican Republic, inclusion of effective therapy for chancroid would be particularly important in the Dominican population studied.
Because 43% of patients with genital herpes would be treated for syphilis and chancroid in the WHO flow chart (Table 5A), we assessed a modified algorithm that withheld therapy for syphilis and chancroid if the clinical examination showed genital vesicular lesions. The resulting sensitivity, specificity, and positive predictive value are shown in Table 5B. For the detection of syphilis or chancroid, the sensitivity, specificity. and positive predictive value of the modified algorithm were 92%, 21%, and 34% in the Dominican Republic and 78%, 28%, and 15% in Peru. The decrease in sensitivity is partly explained by the fact that two patients with vesicles had both HSV and H ducreyi revealed by M-PCR. Overall, the modified algorithm would have assigned overtreatment for chancroid and syphilis to 91% and 82%, respectively, of all subjects in both countries combined and would have prevented administration of antibacterial drugs to 27 patients with genital herpes (n = 22) or no identified pathogen (n = 5). The overall percentage of patients without chancroid or syphilis who would be offered treatment for both of these infections would decrease from 76% with the WHO algorithm to 58% with the modified algorithm. The latter algorithm was more sensitive but less specific than clinical diagnosis for the diagnosis of syphilis and chancroid.
Disease-specific treatment of genital ulcers on the basis of clinical findings alone has low sensitivity, specificity, and predictive value. 6,20 In diverse populations, the sensitivity and specificity of the various WHO algorithms for syndromic management of STD can generally be somewhat improved, and overtreatment reduced, by customizing the risk assessment and laboratory tests used to fit the local population. 21 In areas where there is a high prevalence of HSV, distinguishing between vesicular and nonvesicular lesions on examination may somewhat decrease overtreatment for syphilis and chancroid. In populations with higher prevalences of HIV infection and immunosuppression, even higher rates of HSV might be expected in genital ulcers. 22 The RPR or VDRL was positive at titers ≥1:4, with a confirmatory treponemal test, in 7 of 9 cases of syphilis detected by M-PCR, whereas syphilis was clinically diagnosed in only 4 of 10.
In populations with very low proportions of genital ulcers caused by chancroid (e.g., in the Lima male population studied), it might be reasonable to reserve therapy for chancroid for those who have genital ulcers not associated with vesicular lesions and that persists or progresses longer than the typical natural history of initial episodes of genital herpes. However, from the practical standpoint, the proportion of genital ulcer cases due to chancroid is currently monitored in few if any Latin American settings; the number of clinical settings in which rapid tests for syphilis are performed is quite limited, and follow-up is not certain for many patients. Therefore, syndromic management using the modified algorithm in Table 5B currently is preferable in most clinical settings.
However, because HSV can coexist with other causes of genital ulcers, and other etiologies may occasionally present with vesicular lesions in some patients, patients with vesicular lesions should be observed if antimicrobial treatment is not offered to ensure that the lesions resolve spontaneously and to exclude other etiologies if lesions persist. Of course, in immunosuppressed HIV-infected patients with genital herpes, the lesions might also persist, with or without antibacterial treatment.
Many patients reported antibiotic use for the current episode of GUD before enrollment in the study. It is not known what percentage of treatment was obtained without prescription. The most common treatment reported by the Dominicans was a short course of ampicillin, which would not be highly effective for syphilis and would be useless for HSV, inadequate for donovanosis, and ineffective against β-lactamase-producing strains of H ducreyi. Efforts to eliminate self-administered antibiotic therapy for patients with recurring vesicular lesions, to encourage patients to visit clinicians before beginning therapy, and to encourage pharmacy vendors to refer clients to qualified clinicians and to be aware of syndromic management guidelines may improve the rates of effective antibiotic use for curable forms of GUD.
In this study, the M-PCR test served as the gold standard for the diagnosis of genital ulcer. Previous studies 17,23–26 have found the M-PCR test to be more sensitive than other laboratory tests for the detection of genital ulcer pathogens. Among patients in the Dominican Republic who were T pallidum–negative by PCR in this study, 2 of 77 were dark-field microscopy–positive and 3 of 77 had VDRL titers ≥1:4 and were FTA-ABS-positive. Although these were not considered diagnostic of syphilis according to our gold standard, these laboratory results for patients with genital lesions would certainly merit treatment for syphilis. We cannot rule out the possibility that the dark-field examinations were falsely positive, that VDRL/FTA-ABS titers reflected past disease, or that some M-PCR tests were falsely negative.
The M-PCR test appeared quite sensitive for detection of H ducreyi and HSV. In the Dominican Republic, where culture data were available, the M-PCR test detected 2.6 times as many cases of chancroid as did culture, consistent with the findings of previous work. 17,23,25 One patient was culture-positive but negative for H ducreyi by M-PCR, perhaps indicating problems in sample collection for the PCR sample. For the diagnosis of herpes, M-PCR was more sensitive than Herpchek.
The number of M-PCR samples from Peru that were negative (48%) was relatively large. Some other clinical studies in which the M-PCR assay was used yielded lower negativity rates, 17,23–25 comparable to or even less than the 26% of Dominican Republic samples that were negative in this study. We collected samples in Digene specimen transport medium rather than the now-available Roche specimen transport medium, necessitating the 10-fold dilution of specimens before analysis. This may have increased the negativity rate by M-PCR. Neither PCR inhibitors (which could be detected in this assay) nor the duration of the ulcer could explain the PCR-negative results. However, in spite of the relatively large number of negatives by M-PCR in the Peruvian data set, this test was still the most sensitive laboratory test for detection of genital ulcer pathogens. Several other factors could contribute to negative results in all studies; the most obvious is variation in the inpatient sample selection or specimen collection and handling. Alternatively, pathogens other than T pallidum, H ducreyi, and HSV or physical or chemical damage could be responsible for the negativity of some ulcers by M-PCR in Peru. The ulcers from at least two of the PCR-negative patients from the Dominican Republic were determined to be due to LGV and donovanosis on the basis of the laboratory findings.
In summary, clinical etiologic diagnosis of GUD was inaccurate and unsuitable as a basis for management of genital ulcer lesions. Laboratory tests other than RPR are unavailable in many developing countries, and even in industrialized countries, tests such as dark-field examination and H ducreyi culture are cumbersome, and the latter is insensitive. The M-PCR test is an attractive alternative for the detection of genital ulcer pathogens. For studies in which the presence of genital ulcer pathogens is sought as a guide to syndromic management, the M-PCR test has proven invaluable and is most suitable for collection in field conditions, with M-PCR tests being performed in reference laboratories. Although not yet rapid, affordable, or simple enough to use as a clinical test for case management in developing countries, the test holds promise for industrialized countries. Currently, the M-PCR test is not commercially available. However, the procedure for this test has been published, 26 and such tests could be performed in reference laboratories. For developing countries, syndromic management of genital ulcers is recommended, guided by (1) an algorithm modified to triage patients with and without vesicular lesions to different managements in settings where herpes is common but chancroid uncommon; and (2) where feasible, in Latin American settings where syphilis is uncommon and serologic testing for syphilis is available, consideration of the use of a sensitive serologic test for syphilis to establish the diagnosis and the need for treatment of patients and partners for syphilis. For example, the rapid test could consist of a rapid treponemal antibody dipstick test, followed by confirmation by quantitative VDRL test.
1. World Health Organization/Global Programme on AIDS. Global Prevalences and Incidences of Selected Curable Sexually Transmitted Diseases: Overview and Estimates [WHO/GPA/STD95:1]. Geneva: World Health Organization/Global Programme on AIDS, 1995:1–26.
2. World Health Organization. Sexually Transmitted Disease: Three Hundred and Thirty Three Million New, Curable Cases in 1995 [press release, WHO/64]. Geneva: World Health Organization, 1995.
3. Hayes RJ, Schulz KF, Plummer FA. The cofactor effect of genital ulcers on the per-exposure risk of HIV transmission in sub-Saharan Africa. J Trop Med Hyg 1995; 98: 1–8.
4. Royce FA, Sena A, Cates W, Cohen MS. Sexual transmission of HIV. N Engl J Med 1997; 336: 1072–1078.
5. Jackson DJ, Rakwar JP, Bwayo JJ, Kreiss JK, Moses S. Urethral Trichomonas vaginalis
infection and HIV-1 transmission. Lancet 1997; 350: 1076.
6. Chapel TA, Brown WJ, Jefferies C, Stewart JA. How reliable is the morphologic diagnosis of penile ulcerations? Sex Transm Dis 1977; 4: 150–154.
7. World Health Organization/Global Programme on AIDS. Management of Sexually Transmitted Diseases [WHO/GPA/TEM/94.1]. Geneva: World Health Organization, 1994.
8. World Health Organization. STD Case Management. STD Case Management Workbook 4: Diagnosis and Treatment [WHO/GPA/TCO/PMT/98.18D]. Geneva: World Health Organization, 1998.
9. Costello DC, Maggwa N, Mati JK, Solomon M, Mbugua S, Hunter DJ. Risk Factors for gonorrhea, syphilis, and trichomonas infections among women attending family planning clinics in Nairobi, Kenya. Genitourin Med 1994; 70: 155–161.
10. Brunham RC, Ronald AR. Epidemiology of sexually transmitted diseases in developing countries. In: Wasserheit JB, Aral SO, Holmes KK, eds. Research Issues in Human Behavior and Sexually Transmitted Diseases in the AIDS Era. Washington, DC: American Society for Microbiology, 1991: 61–80.
11. Purevdawa E, Moon TD, Baigalmaa C, Davaajav K, Smith M, Vermund SH. Rise in sexually transmitted diseases during democratization and economic crisis in Mongolia. Int J STD AIDS 1997; 8: 398–401.
12. Van Dyck E, Bogaerts J, Smet H, Tello WM, Mukantabana V, Piot P. Emergence of Haemophilus ducreyi
resistance to trimethoprim-sulfamethoxazole in Rwanda. Antimicrob Agents Chemother 1994; 38: 1647–1648.
13. Brasseur O, Ronald A, Piot P. STD Drugs. In: Dallabetta G, Laga M, Lamptey P, eds. The Control of Sexually Transmitted Diseases: A Handbook for the Design and Management of Programs. Arlington, VA: AIDSCAP/Family Health International, 1996.
14. Knapp JS, Wongba C, Limpakarnjanarat K, et al. Antimicrobial susceptibilities of strains of Neisseria gonorrhoeae
in Bangkok, Thailand: 1994–1995. Sex Transm Dis 1997; 24: 142–148.
15. Aplasca De Los Reyes MR, Pato-Mesola V, Klausner JD, et al. A randomized trial of ciprofloxacin versus cefixime for treatment of gonorrhea after rapid emergence of gonococcal ciprofloxacin resistance in the Philippines. Clin Infect Dis 2001; 32: 1313–1318.
16. Orle KA, Gates CA, Martin DH, Body BA, Weiss JB. Simultaneous PCR detection of Haemophilus ducreyi
, Treponema pallidum
, and herpes simplex virus types 1 and 2 from genital ulcers. J Clin Microbiol 1996; 34: 49–54.
17. Totten P, Stamm W. Clear broth and plate media for culture of Haemophilus ducreyi
. J Clin Microbiol 1994; 32: 2019–2023.
18. Totten PA, Stamm WE, Morse SA. How to detect Haemophilus ducreyi
in genital ulcers and why you should care. Clin Microbiol Newsletter 1997; 18: 145–152.
19. Hook EW, Roddy RE, Lukehart SA, Home J, Holmes KK, Tam MR. Detection of Treponema pallidum
in lesion exudate with a pathogen-specific monoclonal antibody. J Clin Microbiol 1985; 22: 241–244.
20. 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.
21. Mindel A, Dalabetta G, Gerbase A, Holmes KK, eds. Syndromic Approach to STD Management [supplement]. Sex Transm Infect 1998; 74(suppl 1):S1–S178.
22. Wawer MJ, Sewankambo NK, Serwadda D, et al. Control of sexually transmitted diseases for AIDS prevention in Uganda: a randomised community trial. Rakai Project Study Group. Lancet 1999; 353: 525–535.
23. 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. J Infect Dis 1997; 175: 583–589.
24. 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.
25. Mertz KJ, Weiss JB, Webb RM, et al. An investigation of genital ulcers in Jackson, Mississippi, with use of a multiplex polymerase chain reaction assay: high prevalence of chancroid and human immunodeficiency virus infection. J Infect Dis 1998; 178: 1060–1066.
© Copyright 2002 American Sexually Transmitted Diseases Association
26. Orle KA, Weiss JB. Detection of Treponema pallidum, Haemophilus ducreyi,
and herpes simplex virus by multiplex PCR. In: Peeling RW, Sparling PF, eds. Sexually Transmitted Diseases: Methods and Protocols. Totowa, NJ: Human Press, 1999: 67–79.