These authors contributed equally to this work.
Diagnosis of treponemal co-infection in HIV-infected West Africans
Article first published online: 20 SEP 2012
© 2012 Blackwell Publishing Ltd
Tropical Medicine & International Health
Volume 17, Issue 12, pages 1521–1526, December 2012
How to Cite
Mamoojee, Y., Tan, G., Gittins, S., Sarfo, S., Stephenson, L., Carrington, D., Bedu-Addo, G., Phillips, R., Appiah, L. T. and Chadwick, D. (2012), Diagnosis of treponemal co-infection in HIV-infected West Africans. Tropical Medicine & International Health, 17: 1521–1526. doi: 10.1111/j.1365-3156.2012.03090.x
- Issue published online: 21 NOV 2012
- Article first published online: 20 SEP 2012
- sensitivity and specificity
- sensibilité et spécificité
- sensibilidad y especificidad
Objectives To evaluate the performance of two enzyme immunoassays (EIA), Murex and ICE, and the Determine TP point-of-care test (POCT) in diagnosing treponemal infection (syphilis or yaws) in patients attending a large HIV clinic in Ghana; to determine the prevalence of treponemal co-infections; and to characterise demographic and clinical features of patients with infection.
Methods Samples were tested with EIAs and rapid plasma reagin (RPR), then POCT and reference assays for Treponema pallidum to determine prevalence of active and past infection. Sensitivity and specificity of each assay were calculated and demographic and clinical characteristics of patients compared. Data were collected from case notes of patients retrospectively.
Results Overall, 45/284 patient samples (14.8%, 95% CI, 11.1–19.4%) were Treponema pallidum particle agglutination (TPPA) positive, and of these, 27 (64.3%) were RPR positive and 4 (8.9%) were treponemal IgM positive. Both EIAs and Determine TP POCT showed high sensitivities and specificities for identifying infection although RPR was less reliable. Clinical features of syphilis or yaws were rarely identified in TPPA-positive patients suggesting most had previous or late latent infection. Treatment of various intercurrent infections using short courses of antibiotics active against T. pallidum was common in the clinic.
Conclusions A high proportion of this HIV-infected cohort showed evidence of treponemal infection. Both EIAs as well as the POCT were practical and effective at diagnosing treponemal co-infection in this setting. RPR alone was unreliable at identifying active treponemal co-infection, however might be useful in some settings where treponemal-specific assays are unaffordable.
Objectifs: Evaluer la performance de deux tests immunoenzymatiques (EIA), Murex et ICE et le test ‘Determine TP point-of-care (POCT)’ dans le diagnostic de l’infection tréponémique (syphilis ou pian) chez les patients fréquentant une grande clinique du VIH au Ghana; déterminer la prévalence de la coinfection tréponémique et caractériser les caractéristiques démographiques et cliniques des patients infectés.
Méthodes: Les échantillons ont été testés avec les tests EIA et le test rapide de la réagine plasmatique (RPR), puis avec le test POCT et le test de référence pour T. pallidum (TPPA), afin de déterminer la prévalence de l’infection active et précédente. La sensibilité et la spécificité de chaque test ont été calculées et les caractéristiques démographiques et cliniques des patients ont été comparées. Les données ont été collectées rétrospectivement à partir des notes de cas sur les patients.
Résultats: Dans l’ensemble, 45/284 échantillons de patients (14.8%, IC95%: 11.1–19.4%) étaient positifs pour TPPA et de ceux-ci, 27 (64.3%) étaient RPR positifs et 4 (8.9%) étaient positifs pour les IgM tréponémiques. Les tests EIA et POCT ont tous deux démontré des sensibilités et des spécificités élevées pour identifier l’infection, alors que le test RPR était moins fiable. Les caractéristiques cliniques de la syphilis ou pian ont rarement été identifiées chez les patients positifs pour TPPA, suggérant que la plupart avaient des infections précédentes ou latentes tardives. Des traitements d’infections intercurrentes diverses utilisant des régimes d’antibiotiques de courte durée, actifs contre T. pallidumétait courants dans la clinique.
Conclusions: Une forte proportion de cette cohorte infectée par le VIH a montré des signes d’infection tréponémique. Les tests EIA ainsi que le test POCT étaient pratiques et efficaces pour diagnostiquer la coinfection tréponémique dans ce contexte. Le test RPR seul n’était pas fiable pour identifier les coinfections tréponémiques actives, mais pourrait être utile dans certains contextes où des tests tréponémiques spécifiques sont inabordables.
Objetivos: Evaluar el desempeño de dos inmunoensayos enzimáticos (IEE), Murex e ICE, y la prueba rápida in situ (POCT) Determine TP en el diagnóstico de infección por treponema (sífilis o pián), en pacientes de una gran clínica para VIH en Ghana; determinar la prevalencia de coinfecciones con treponema; y determinar las características demográficas y clínicas de los pacientes con infección.
Métodos: Con las muestras se realizaron las pruebas de IEEs y la Prueba Rápida de Reagina (RPR), después la POCT y pruebas de referencia para T. pallidum, con el fin de determinar la prevalencia de infecciones activas y pasadas. Se calcularon la sensibilidad y la especificidad de cada ensayo y se compararon las características clínicas y demográficas de los pacientes. Los datos se recogieron de forma retrospectiva de las historias clínicas de los pacientes.
Resultados: En general, 45/284 muestras de pacientes (14.8%, 95% IC, 11.1–19.4%) eran TP-PA positivas, y de estas 27 (64.3%) eran RPR positivas y 4 (8.9%) eran positivas para IgM de treponema. Tanto las IEEs como las POCT Determine TP mostraron una alta sensibilidad y especificidad para identificar la infección, aunque la RPR eran menos confiable. Las características clínicas de la sífilis o el pián eran rara vez identificadas en pacientes TP-PA positivos, sugiriendo que la mayoría habían tenido infecciones previas o latente tardías. El tratamiento de varias infecciones concurrentes, utilizando tratamientos cortos de antibióticos activos frente a T. pallidum, era una práctica común en la clínica.
Conclusiones: Una alta proporción de esta cohorte infectada con VIH mostró evidencia de infección por treponema. Tanto las IEEs como las POCT eran prácticas y efectivas en el diagnóstico de una coinfección con treponema, aunque podría ser útil en emplazamientos en los que las pruebas específicas para treponema son inasequibles.
Both syphilis and HIV infections are prevalent in Africa (Assefa et al. 1994; Cossa et al. 1994; Killewo et al. 1994), whilst yaws is also prevalent in many tropical areas of Africa including Ghana (Asiedu 2008). Syphilis may present atypically in HIV infection, progresses more rapidly to tertiary syphilis (Johns et al. 1987; Musher et al. 1990; Hutchinson et al. 1994), is associated with impaired responses to therapy (Musher et al. 1990) and poses an increased risk of mother to child transmission of HIV in co-infected women (Mwapasa et al. 2006). Furthermore, HIV co-infected patients are more likely to present with secondary (as opposed to primary) syphilis or persistent chancres (Hutchinson et al. 1994). Little is known about the effect of HIV infection on the natural history of yaws. Diagnosing and treating syphilis co-infection not only reduces complications of the disease but also reduces rates of transmission of both syphilis and HIV (Wasserheit 1992; Del Mar Pujades et al. 2002; Kapiga et al. 2007); however, few HIV clinics in sub-Saharan Africa routinely test patients for syphilis. The prevalence of syphilis in HIV-infected populations is not known in many parts of Africa, although estimates in pregnant women range from <1% to 8% (Sombie et al. 2000; Mwapasa et al. 2006; Urassa et al. 2001); studies in Kenya and Nigeria found rates of around 2–3% in both men and women with HIV infection (Guthrie et al. 2009; Nnoruka et al. 2005).
Diagnosing syphilis or yaws through serological assays presents various difficulties in HIV infection. Non-treponemal-specific tests, such as Venereal Disease Research Laboratory (VDRL) and rapid plasma reagin (RPR), may be falsely positive or negative, in some cases because of the prozone phenomenon, an effect where false negative results occur in samples that are undiluted or at low dilution which become positive in the same test at higher dilutions of the sample. The phenomenon is attributed to an interference in the normal binding of antigenic and antibody interactions, which happens when high antibody concentrations compete with a disproportionately lower amount of antigen than is optimal for binding of the antigen–antibody complex (Taniguchi et al. 1995; Smith & Holman 2004). Treponemal-specific tests, such as enzyme immunoassays (EIA), Treponema pallidum haemagglutination assay (TPHA) and fluorescent treponemal antibody using absorbed sera (FTA-Abs), are more specific tests, but may yield false negative results in HIV co-infection (Erbelding et al. 1997; Sordillo et al. 1998). Furthermore, these assays do not readily distinguish between current or previous infection with T. pallidum and cannot discriminate between venereal and non-venereal treponemal infections including yaws (Treponema pertenue), which is endemic in some African countries. There are also considerable technical challenges in performing some of these assays in developing world settings.
In Ghana, the prevalence of HIV infection in the general adult population is estimated to be 1.9% (UNAIDS 2008 update). The prevalence of syphilis ranges between 5.6% among antenatal clinic attendants and 17% in prison inmates (Apea-Kubi et al. 2004; Adjei et al. 2006). The prevalence of syphilis in the HIV-infected population in Ghana is largely unknown, and routine testing for syphilis is not currently included in standard HIV care programmes. As antiretroviral therapy (ART) programmes have been established in Africa with encouraging results in many countries (May et al. 2010), establishing the presence of co-infections such as syphilis becomes more important, particularly where treatment is relatively simple and cost-effective. The objectives of this study were to estimate the seroprevalence of syphilis or yaws within an HIV-infected outpatient population in Ghana, to characterise clinical features in patients with positive serology and assess the performance of three different treponemal-specific assays in this patient group and setting.
Patients and data
We conducted a cross-sectional study on HIV-infected adult patients attending a government hospital in Kumasi, Ghana between 2007 and 2009. Stored serum and plasma samples were identified and tested, and anonymised data retrospectively extracted from case notes of randomly selected patients attending the Komfo Anokye Teaching Hospital HIV clinic in Kumasi. The study was approved by the Committee on Human Research and Ethics at KNUST (Kumasi, Ghana). Patients attending the clinic were almost entirely over the age of 18, and the majority were taking ART. Pregnant women did not normally attend this clinic, and patients were generally reviewed every 3 months, unless they had high CD4 counts and were not on therapy, in which case, they were reviewed every 6 months. Demographic and clinical data were collected and included information from both before the samples were obtained and up to 3 years afterwards.
The following assays were performed in Kumasi, according to the manufacturer’s instructions. Murex Syphilis and ICE syphilis IgG + IgM ELISA (Murex Biotech Limited, Dartford, UK) assays were undertaken using a standard plate reader. An RPR assay (Abbott Diagnostics, UK) using cards was performed, with serial dilutions for all positive tests and a number of samples with a negative tests also evaluated to exclude the prozone effect. Remaining samples were shipped to Bristol (HPA Treponemal Reference Laboratory, UK) for further testing using the following assays. Treponema pallidum particle agglutination (TPPA) screening assay (Fujirebio Inc.) was performed according to the manufacturer’s instructions, and sample was diluted 1:2–1:80. RPR test – Axis Shield RPR (Lab21) assay was performed according to the manufacturer’s instructions on all samples testing positive previously, with serial dilutions to establish titre. Where there was a discrepant result with the RPR test in Ghana, the test was repeated a second time and this result used as the final arbiter. Newmarket Syphilis EIA II (Lab21) was performed according to the manufacturer’s instructions on all samples with equivocal TPPA result or negative TPPA result and positive EIA. Mikrogen Recomwell Treponema IgM Test (Mikrogen Diagnostics GmbH, Germany) was performed according to the manufacturer’s instructions on all TPPA-positive or equivocal samples. Treponema pallidum IgM Marblot Strip test system (MarDx; Trinity Biotech) was performed according to the manufacturer’s instructions only on samples positive for IgM and discordant for TPPA and EIA tests. Inno-Lia Syphilis Score (Microgen) was performed according to the manufacturer’s instructions only on samples positive for TPPA and negative on the EIAs. Determine Syphilis TP (Alere) was performed according to the manufacturer’s instructions on patients with sufficient sample left after the other assays.
Interpretation of assay results
A confirmed positive result, indicative of previous or current treponemal infection, was defined as a confirmed positive TPPA or positive Newmarket EIA and equivocal TPPA. A confirmed negative result was a negative for both tests. All other result combinations were analysed with the negative results. Patients with confirmed positive TPPA and positive RPR were defined as having active or latent (untreated) treponemal infection and those with negative RPRs as having either previously treated (inactive) infection or latent (false negative RPR) infection, except for patients with positive IgM, who were considered to have early infection. Patients with negative TPPA and positive IgM were assumed to have very recent infection or primary syphilis.
Statistical analysis was conducted using the chi-squared and Student’s t-tests to assess the significance of any difference in demographics and clinical parameters between the two groups. Sensitivity, specificity and predictive values of each assay were calculated for TPPA-positive patients using standard methods for proportions and exact 95% confidence limits determined.
Samples from 284 patients were obtained, and there was sufficient to test all samples by the ICE and MUREX EIAs and reference assays (TPPA, +/− Newmarket EIA and IgM). There was only sufficient sample to test 192 patients’ samples using the Determine TP test. It was possible to obtain complete demographic and clinical data corresponding to these samples from case notes of 228 patients, which are shown in Table 1. The patient characteristics of the group sampled were similar to the general characteristics of the HIV clinic population and were similar between TPPA-positive and negative groups in relation to demographics, CD4 counts and HIV (WHO) clinical stage. TPPA-positive patients showed a trend towards having lower body mass indices, being more often single (rather than married) and being less educated. Clinical features consistent with active treponemal disease were not frequently seen in either group. Non-specific skin rashes, lymphadenopathy and features of peripheral neuropathy were noted in approximately 16%, 2% and 9% of the TPPA-positive group and 11%, 1% and 9% of the TPPA-negative group. Most of the neurological symptoms were attributed to either current or previous antiretroviral drugs, HIV or herpes zoster infections. The only significant difference between groups was noted in the number of patients reporting current symptoms (or recent treatment) of sexually transmitted infection: 9% of the TPPA-positive group and 2% of the TPPA-negative group (P = 0.01). Only one of the four treponemal IgM-positive patients had a clinical feature consistent with early syphilis, in this case a generalised rash. A significant number of TPPA-positive patients (23%) had been treated at some point in the clinic with short courses of oral antibiotics against T. pallidum, usually for intercurrent respiratory tract or urinary tract infections. There were no instances of patients suspected of having syphilis or being treated for syphilis in the clinical notes.
|TPPA-positive patients (n = 43)||TPPA-negative patients (n = 185)||P|
|Age (SD)||41.2 (12.2)||39.5 (8.7)||0.22|
|Gender (% female)||67||73||0.45|
|BMI, kg/m2 (SD)*||22.4 (4.2)||24.5 (4.6)||0.03|
|CD4 count, cells/μl (SD)*||417 (261)||420 (162)||0.95|
|CD4 count at diagnosis, cells/μl (SD)||321 (325)||270 (522)||0.42|
|WHO stage, %|
|On antiretroviral therapy (%)||79||81||0.83|
|Marital status (% married)||35||51||0.07|
|Educational attainment, %†|
Serological test results are shown in Figure 1. In summary, 42/284 patient samples (14.8%, 95% CI, 11.1–19.4%) were TPPA positive, and of these, 6/42 (14.3%) were treponemal IgM positive, indicating recent treponemal infection in this population. A further three TPPA-negative samples were positive for IgM, although not confirmed by Western blot. A significant proportion (around 20%) of RPR results were discrepant between those performed in Ghana and those in the reference laboratory, mostly false positives in Ghana. Of the TPPA-positive samples, 27/42 (64.3%, 95% CI, 49.2–77.0%) were RPR positive confirming active treponemal disease. Of these, all were positive on neat serum. No samples (using either RPR assay) were negative on neat serum but positive after dilution, that is, none illustrating the prozone phenomenon. Twenty-five patients’ samples (8.8%, 95% CI, 6.0–12.7%) gave false-positive RPR tests, that is, RPR only and negative in all other treponemal-specific tests.
Of the samples tested using the ICE and MUREX EIAs and RPR, 266 patients had interpretable results and 192 Determine TP tests were analysed. Sensitivities, specificities and predictive values for the assays are shown in Table 2. All three treponemal-specific assays showed good sensitivities and specificities, with the Determine TP having the best overall performance, whilst the ICE EIA compromised sensitivity slightly for higher specificity. As a single test, RPR showed the lowest sensitivity and specificity of all tests performed.
|Treponemal Assay||Sensitivity % (95% CI)||Specificity % (95% CI)||Positive Predictive Value % (95% CI)||Negative Predictive Value % (95% CI)|
|Murex EIA||100.00 (91.43–100.00)||96.54 (93.04–98.38)||83.67 (70.96–91.48)||100.00 (98.31–100.00)|
|ICE EIA||92.31 (79.68–97.35)||98.73 (96.33–99.57)||92.31 (79.68–97.35)||98.73 (96.33–99.57)|
|Determine TP*||100.00 (91.62–100.00)||97.33 (93.34–98.96)||91.30 (79.67–96.56)||100.00 (97.44–100.00)|
|RPR||64.10 (48.42–77.26)||89.63 (85.14–92.88)||62.50 (47.03–75.78)||98.63 (96.05–99.53)|
The seroprevalence of treponemal infection among the adult HIV population of Kumasi was estimated to be 15%, which is higher than published prevalence data among blood donors and pregnant women in Ghana, and higher than many HIV-infected populations elsewhere in Africa (Sombie et al. 2000; Mwapasa et al. 2006; Urassa et al. 2001; Guthrie et al. 2009). A significant proportion of this prevalence may represent previous exposure to yaws in the local population because yaws, like syphilis, will give a positive TPPA result. Even if patients with positive TPPA and negative RPR results are presumed largely to have previous, rather than active infection, a prevalence of active treponemal disease of 10% in our cohort is high and indicates a significant disease burden. We believe most active treponemal infection in this population is caused by syphilis for several reasons. First, yaws largely affects children (mean age, 10; Rinaldi 2008), and the mean age of the TPPA-positive group was 41 years. Second, very few clinical signs consistent with yaws were recorded in the medical notes of those patients with positive TPPA and RPR. Third, HIV-infected individuals are more likely than the general population to contract other sexually transmitted infections. Hence, given the likely high seroprevalence of syphilis in this population, and the simplicity of treating it, there is a strong argument to incorporate syphilis screening into HIV care programmes in Ghana.
Our study showed that patients with treponemal infection were more likely to be unmarried, have lower educational attainment and have recent sexually transmitted infections, which is consistent with data from other studies. Our observation that a high proportion of patients with active treponemal disease received short courses of oral antibiotics with activity against T. pallidum is worrying because partial treatment of syphilis may modify its subsequent presentation and lead to delayed or missed diagnosis of late-stage disease. Clearly given the retrospective nature of this study, there are considerable limitations in interpretation of clinical data from the case notes review. There was, however, little indication of any specific features of late-stage syphilis in the TPPA-positive patients nor any clear features of early syphilis in most of the IgM-positive patients. A prospective study including dark field microscopy of patients presenting with possible chancres (or other features of early syphilis), as well as serial serology, would provide better data to determine whether patients were largely presenting with syphilis, as opposed to non-venereal treponemal infections, as we suspect.
Both the ICE EIA and Determine TP point-of-care test (POCT) performed well in patients infected with HIV in terms of sensitivity and specificity; however, in the field, RPR was less reliable with relatively frequent false positives, which may have occurred because of sample degradation during storage, whilst the RPR assay used in the reference laboratory produced more false negatives. Given that the treatment of syphilis is inexpensive and usually safe, and the potential complications of untreated syphilis are serious, there is a strong case for using an assay with high sensitivity (and potentially lower specificity) in situations where resources and laboratory infrastructure are most limited. Unfortunately, RPR would not appear to fulfil these criteria given the unreliable results produced by both assays we used, and more research is needed either to establish the optimal way the RPR assay is used in this population (ideally for use on fresh serum) or to evaluate other simple and cheap treponemal-specific or non-specific assays that can detect syphilis co-infection with high sensitivity. Moreover, given the possibility of patients with late or tertiary syphilis having negative RPR results, there is a good case for using treponemal-specific assays as first line tests for HIV-infected patients.
In recent years, the development of automated EIAs and chemiluminescence immunoassays (CIA) has prompted the use of reverse sequence syphilis screening in some developed countries with access to this technology (Park et al. 2011; Radolf et al. 2011). This method uses initial screening of samples with treponemal EIA or CIA followed by a non-treponemal assay in reactive samples to indicate active disease. TPPA should be used where resources are available to confirm reactive results (whether EIA only or RPR positive) as TPPA provides the highest specificity and will identify any false-positive results. In developing countries such as Ghana, where resources and laboratory infrastructure are less limited than in other parts of Africa, any of the three treponemal-specific assays we evaluated could be used initially to confirm syphilis co-infection. Although slightly more expensive, the Determine TP assay has the advantage of being a rapid POCT which, like RPR, does not require any laboratory equipment. However, the EIAs might be more suitable for clinics such as that in Kumasi where large numbers of patients attend and there is access to a EIA plate reader for screening large numbers of samples. In developing countries that are more resource limited, using RPR assays alone to screen for syphilis co-infection in HIV clinics will lead to a significant number of patients with treponemal infection who are not identified and some patients with false-positive results who do not have active syphilis being inadvertently treated. However, this may be preferable to no screening at all.
Enzyme immunoassays and Determine kits were provided by Abbott Diagnostics/Murex Biotech Ltd. We are grateful to the patients and staff at KATH Department of Medicine for their contribution to this study and Rachel Turnbull for providing technical assistance and advice with assays. This study was funded through the South Tees Infectious Diseases Research Fund.
- 2006) Prevalence of human immunodeficiency virus, hepatitis B virus, hepatitis C virus and syphilis among prison inmates and officers at Nsawam and Accra, Ghana. Journal of Medical Microbiology55, 593–597. , , et al. (
- 2004) Neisseria gonorrhoea, Chlamydia trachomatis and Treponema pallidum infection in antenatal and gynecological patients at Korle-Bu Teaching Hospital, Ghana. Japanese Journal of Infectious Diseases57, 253–256. , , , , & (
- 2008) The return of yaws. Bulletin of the World Health Organization86, 507–508. (
- 1994) Seroprevalence of HIV-1 and syphilis antibodies in blood donors in Gonder, Ethiopia, 1989–1993. Journal of Acquired Immune Deficiency Syndromes7, 1282–1285. , , & (
- 1994) Syphilis and HIV infection among displaced pregnant women in rural Mozambique. International Journal of STD and AIDS5, 117–123. , , et al. (
- 2002) Herpes simplex virus type 2 infection increases HIV incidence: a prospective study in rural Tanzania. AIDS16, 451–462. , et al. (
- 1997) Syphilis serology in human immunodeficiency virus infection: evidence for false-negative fluorescent treponemal testing. Journal of Infectious Diseases176, 1397–1400. , , et al. (
- 2009) Sexually transmitted infections among HIV-1-discordant couples. PLoS ONE4, e8276. , , , , et al. (
- 1994) Altered clinical presentation of early syphilis in patients with human immunodeficiency virus infection. Annals of Internal Medicine121, 94–100. , , , & (
- 1987) Alteration in the natural history of neurosyphilis by concurrent infection with the human immunodeficiency virus. New England Journal of Medicine316, 1569–1572. , & (
- 2007) The role of herpes simplex virus type 2 and other genital infections in the acquisition of HIV-1 among high-risk women in northern Tanzania. Journal of Infectious Diseases195, 1260–1269. , , et al. (
- 1994) Prevalence and incidence of syphilis and its association with HIV-1 infection in a population-based study in the Kagera region of Tanzania. International Journal of STD and AIDS5, 424–431. , , , , & (
- 2010) Prognosis of patients with HIV-1 infection starting antiretroviral therapy in sub-Saharan Africa: a collaborative analysis of scale-up programmes. Lancet376, 449–457. , , et al. (
- 1990) Effect of human immunodeficiency virus (HIV) infection on the course of syphilis and on the response to treatment. Annals of Internal Medicine113, 872–881. , & (
- 2006) Maternal syphilis infection is associated with increased risk of mother-to-child transmission of HIV in Malawi. AIDS20, 1869–1877. , , et al. (
- 2005) Evaluation of syphilis in patients with HIV infection in Nigeria. Tropical Medicine and International Health10, 58–64. & (
- 2011) Screening for syphilis with the treponemal immunoassay: analysis of discordant serology results and implications for clinical management. Journal of Infectious Diseases204, 1297–1304. , , , , & (
- 2011) Discordant results from reverse sequence syphilis screening – five laboratories, United States, 2006–2010. MMWR Morbidity and mortality weekly report60, 133–137. , , et al. (
- 2008) Yaws: a second (and maybe last?) chance for eradication. PLoS Neglected Tropical Diseases2, e275. (
- 2004) The prozone phenomenon with syphilis and HIV-1 co-infection. Southern Medical Journal97, 379–382. & (
- 2000) Seroprevalence of syphilis among women attending urban antenatal clinics in Burkina Faso, The DITRAME Study Group. DIminunation de la TRAnsmission Mere-Enfant. Sexually Transmitted Infections76, 314–316. , , et al. (
- 1998) False-negative fluorescent treponemal tests and confirmation of syphilis infection. Journal of Infectious Diseases178, 294–295. , & (
- 1995) The prozone phenomenon in secondary syphilis. Acta Dermato-Venereologica75, 153–154. , & (
- UNAIDS/WHO. (2008) Epidemiological Fact Sheets on HIV and AIDS, 2008 Update. http://apps.who.int/globalatlas/predefinedReports/EFS2008/full/EFS2008_GH.pdf (accessed 28/02/2012).
- 2001) Risk factors for syphilis among HIV-1 infected pregnant women in Dar es Salaam, Tanzania. African Journal of Reproductive Health5, 54–62. , , , , & (
- 1992) Epidemiolgy synergy: interrelationships between HIV and other STDs. Sexually Transmitted Diseases19, 61–77. (