SEARCH

SEARCH BY CITATION

Keywords:

  • HIV infection;
  • sexually transmitted infections;
  • female sex workers;
  • condom;
  • Madagascar

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

In a cross-sectional study in 1998 we assessed human immunodeficiency virus (HIV) and syphilis infections and their risk factors among the 316 registered female sex workers (FSWs) of Toliary, south-west Madagascar. No case of HIV infection was detected, but 18.4% of registered FSWs had syphilis. Only half of these women regularly used condoms. In a multiple logistic regression analysis, risk factors for syphilis infection were multiple clients per week and, paradoxically, regular use of condoms. The variables associated with irregular use of condoms were younger ages of registered FSWs, multiple clients per week and Malagasy clients. The high prevalence of syphilis infection associated with irregular use of condoms might facilitate a very fast spread of HIV infection among these FSWs. Promotion of condom use and surveillance of sexually transmitted infections and HIV infection incidence are needed in the south of Madagascar.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

Madagascar is an African country with a low incidence of human immunodeficiency virus (HIV) infection in the adult population. In 2000 it was 0.15%, whereas in neighbouring continental areas, such as Mozambique, Zimbabwe, Swaziland or South Africa, HIV prevalence ranges from 13.22% to 25.25% (UNAIDS 2000). Madagascar's insular location, the small number of national and international migrants, and a former national policy of isolation were the most common hypotheses for this prevalence rate. However, present conditions such as increasing sexual tourism, prostitution and a high prevalence of sexually transmitted infections (STIs) could facilitate the spread of HIV (Ravaoarimalala et al. 1998). According to the Madagascar Ministry of Health, the estimated HIV prevalence for 2015 will range from 3% to 15% in the sexually active population (Behets et al. 1996).

We undertook a cross-sectional study in the deprived south-west capital Toliary to determine the prevalence of HIV antibodies and syphilis infection as well as associated risk factors. The investigation was conducted among female sex workers (FSWs), who represent a population that is highly exposed to HIV in Africa (Estébanez et al. 1993).

Study population

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

The study was conducted at the HIV/STI Dispensary of Toliary, where registered FSWs were legally followed every 2 months. Depending on available resources (reagents, personals and drugs), symptomatic women were screened for syphilis and trichomoniasis, and/or underwent syndromic treatment for these two STIs and for gonorrhoea. All registered FSWs who attended between 1 July, and 31 October, 1998, were potentially included in the study. Nurses of the HIV/STI Dispensary provided them with a careful detailed explanation of the purpose and the procedure of this study and a guarantee of their anonymity. After obtaining informed consent, a structured questionnaire was completed for information on sociodemographic characteristics and medical history of STIs. The FSWs were asked to self-report current genital symptoms, and to state if they had been diagnosed with an STI at the time. Pre- and post-test counselling were available using routine means.

Laboratory exams

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

A 7-ml blood sample was collected in the HIV/STI Dispensary from each participating registered FSW, and immediately sent to the regional STI reference laboratory for processing. An RPR test (Redi Test®, Biokit S.A., Barcelona, Spain) was conducted to assess the seroprevalence of syphilitic markers. The WHO method of anonymous and unlinked screening was performed for HIV antibody detection (CHIN 1990), using an enzyme immunoassay (EIA) (Génélavia Mixt®, Sanofi Diagnostics Pasteur, Paris, France), and positive or equivocal results were re-examined with a second EIA (HIV Spot®, made by E.Y-LABS for Grenelabs Diagnostics, Singapore). In case of discordant results with the two EIAs, the sera were sent to the Institut Pasteur of Madagascar for alternative EIA and/or Western blotting. To respect Malagasy legislation, all HIV-positive or undetermined sera were referred to the HIV Reference National Laboratory (LNR) in the capital, Antananarivo. None of the documents (questionnaires or laboratory records) contained the women's names, preventing matching of biological results with their identities.

Statistical analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

Chi-square and Fisher's exact tests were used for comparison of proportions, and multiple logistic regression was performed to identify independent factors associated with the declaration of irregular condom use and a positive RPR test result. Variables were included in the model at P < 0.10 and removed at P > 0.15. P-values under 0.05 were considered significant. Statistical analysis was undertaken with SPSS 7.5® (Chicago, IL, USA).

Ethical authorizations

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

This study was approved by the ‘Plan National de Lutte contre le SIDA’ and the Regional Public Health authority (no National Ethical Committee was functional at these times). It was conducted under the responsibility of the Director of the Direction Inter Régionale de Développement Sanitaire of the province of Toliary.

Study population

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

All 316 registered FSWs who attended the dispensary during the test period agreed to participate. Table 1 reports their demographic and behavioural characteristics. The median age was 23 years (range 15–42), and median age at first sexual intercourse was 17 years (11–22). The average number of children was 0.9 (0–6). Sixty-seven per cent of the registered FSWs had sexual intercourse only with Malagasy clients, 24.9% with Malagasy or foreign clients, 4.4% strictly with foreigners, and 3.7% did not answer the question. The median weekly number of clients was 2 (1–40), and the median of sexual encounters was 2 per week (1–40). At the time of data collection, 70% of the registered FSWs claimed to be currently suffering from at least one STI previously diagnosed by a health care professional, and 82% of them reported at least one non-specific clinical sign of STI.

Table 1.  Demographic and behavioural characteristics of registered female sex workers in Toliary, Madagascar in 1998
Characteristicsn = 316 max.%
Age
 <205619.0
 20–228227.9
 23–258127.6
 ≥267525.5
Age of first sexual intercourse
 ≤156622.4
 165418.4
 177625.9
 ≥189833.3
Number of children
 013646.9
 18730.0
 ≥26723.1
Frequency of condom use
 Never9733.1
 Sometimes4916.7
 Always14750.2
Number of clients per week
 17626.0
 213947.6
 >27726.4
Number of sexual encounters with clients per week
 1113.8
 212743.5
 38729.8
 ≥46722.9
Self reported non-specific features of STI
 Vaginal discharge14449.0
 Pelvic pruritus12843.5
 Pelvic pain5518.7

Only 50% of the women declared regular use of condoms. But 45.5% of condom users stated that they would not use one if the client requested it, and 30.3% reported problems with using a condom. Ten per cent found it difficult to purchase condoms, mainly because they lived outside the town. The variables associated with absent or partial condom use are reported in Table 2: multiple logistic regression analysis indicated that younger age of the registered FSWs, only one client per week, having sexual intercourse strictly with Malagasy clients and the absence of self-reported STIs were independently associated with non-constant condom use.

Table 2.  Univariate and multivariate associations of demographic and behavioural variables with the absence of or partial condom use among registered female sex workers in Toliary, Madagascar in 1998
Characteristicsn = 316 max.%OR95% CIAdjusted OR95% CI
  • *

     Irregular condom use, i.e no or partial condom use.

  • OR: Odds ratio. 95% CI: 95% confidence interval. OR from multivariate logistic regression adjusting for all variables listed

  • Model characteristics: 2 likelihood = 272.961, Chi-square = 121.849, 8 d.f., P < 0.001.

Irregular condom use*
 Age (years)
  <205687.51.0 1.0 
  20–228252.40.160.06–0.390.120.04–0.34
  23–258030.00.060.02–0.150.060.02–0.18
  ≥267540.00.100.04–0.240.070.03–0.21
 Number of children
  013560.71.0   
  18739.10.410.24–0.72  
  ≥26738.80.410.22–0.75  
 No. of clients per week
  17677.61.0 1.0 
  213934.50.150.08–0.290.130.06–0.28
  >27648.70.270.14–0.550.470.20–1.08
 Origin of clients
  Strictly Malagasy19658.71.0 1.0 
  Foreigners and Malagasy9631.30.320.19–0.540.270.13–0.54
 Self-reported STI (S-STI)
  S-STI–8873.91.0 1.0 
  S-STI+20439.20.230.13–0.400.300.15–0.60

Univariate logistic regressions (Table 3) were also used to assess the specific associations between the frequency of condom use and being RPR+, self-reporting an STI and reporting a non-specific clinical feature of STI. These analyses revealed a positive association of regular condom use with the markers of STI, compared with declared non-use of condoms.

Table 3.  Univariate associations of frequency of condom use with RPR+ test, self-reported sexually transmitted infections (S-STI) and reported non-specific clinical features of STI in Toliary, Madagascar in 1998
  RPR+S-STI+Non-specific self reported clinical features of STI
Condom usen = 316 max.%OR95% CI%OR95% CI%OR95% CI
  1. OR: Odds ratio; 95% CI: 95% confidence interval.

Never9711.31.0 51.01.0 80.41.0 
Sometimes4916.31.50.57–4.163.31.70.82–3.365.30.460.21–0.99
Always14723.82.41.2–5.184.45.22.8–9.489.12.01.0–4.1

Seroprevalence of HIV infection and syphilis

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

None of the 316 registered FSWs presented HIV antibodies. Fifty-four of 294 sera tested for syphilis (18.4%) were positive on RPR testing. None of the registered FSWs reported a syphilitic ulceration, which are frequently unnoticed in women. Table 4 presents the characteristics of these women according to RPR status. The variables independently associated with RPR+ testing were more than two clients a week and regular condom use.

Table 4.  Univariate and multivariate associations of demographic and behavioural variables with RPR-positivity among registered female sex workers in Toliary, Madagascar 1998
Characteristicsn = 316 max.%OR95% CIAdjusted OR95% CI
  1. OR: Odds ratio, 95% CI: 95% confidence interval.

  2. OR from multivariate logistic regression model adjusting for all variables listed.

  3. Model characteristics: 2 likelihood = 264.947, Chi-square = 11.764, 3 d.f., P = 0.008.

Age (years)
 <205614.31.00   
 20–228214.61.030.40–2.71  
 23–258123.51.840.74–4.56  
 ≥267520.01.500.59–3.83  
Number of children
 013617.61.00   
 18717.20.970.48–1.98  
 ≥26722.41.350.65–2.78  
No. of clients per week
 17610.51.00 1.00 
 213917.31.780.76–4.171.310.53–3.24
 >27728.63.401.41–8.232.761.11–6.86
Origin of clients
 Strictly Malagasy
 Foreigners and Malagasy9723.71.650.90–3.03  
Frequency of condom use
 Always14723.81.00 1.00 
 Sometimes or never14613.00.480.26–0.880.530.27–1.02
Condom use
 No difficulty for using20419.1    
 Difficulty for using8915.70.770.40–1.50  
 Non-influenced by clients14822.3    
 Non-use due to clients14514.50.590.32–1.08  

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

The main objectives of this study were to estimate the prevalence of HIV infection in a high-risk group in Toliary, Madagascar, and to assess risk factors that could facilitate the spread of HIV infection. The results were puzzling, with the absence of any case of HIV seropositivity in this high-risk group that contrasts with the presence of high prevalence of syphilis infection and of behavioural risk factors favourable to HIV and STI epidemics.

We have to address some limitations of our study. Socio-economic factors were not assessed, and neither was their potential confounding effect, as low socio-economic situation is a risk factor for HIV and STIs (Lamptey 2002) that can induce women into the commercial sex industry (Ohshige et al. 2000). Other potential confounding factors, such as the ethnic group, also need to be investigated.

HIV infection could have been under detected, as viral load and p24 antigenaemia were not determined. We did not explore potential immune and/or genetic protection against HIV infection among Malagasy (Bienzle et al. 2000), such as specific cytotoxic T-cell anti-HIV response (Rowland-Jones et al. 1993), chemokine receptor CCR5 genotype (Samson et al. 1996), and anti-CCR5 antibodies (Lopalco et al. 2000) because these biological investigations were unavailable to us. We also have to consider the possible presence in this region of a low infectious viral strain, as in Senegal (Kanki et al. 1999). Concerning the biological methodology for syphilis, each RPR+ test should have been confirmed by treponemic reaction [treponema pallidum hemagglutination assay (TPHA) or fluorescent treponema assay absorbant (FTA Abs)] (Young 1992) and if one must use only one test for syphilis screening in an epidemiological survey, the use of the more specific TPHA test instead of a more sensitive cardiololipin-based RPR test would have been more adequate. Hence we may have overestimated the real prevalence of past or active syphilitic infection.

Our work provides insights into some regional characteristics of Madagascar, especially in the deprived region of Toliary, where few studies have been performed because of limited resources. The use of a method based on the resources of an ongoing programme could result in bias, preventing the objectivity required in analytical surveys. We minimized this bias by implementing quality controls during the study, especially for the submission of the questionnaire and the biological analyses. Our study contributed to the improvement of the ongoing programme, transferring new tools to the professionals. Such studies can help motivate health care workers in charge of HIV/STIs control.

The absence of HIV infection detected in this high risk group confirms the very low HIV prevalence in Madagascar, 11 years after the first reported case of HIV seropositivity from a Malagasy FSW in a harbour site (Genin & Coulanges 1988). Why does Madagascar seem to be protected against the HIV pandemic? According to Meda (1995), the country's insular location combined with low levels of national and international migration (road transport is almost inexistent) are possible explanations. Homosexuality and intravenous drug use seems to be rare. However, numerous risk factors, such as high STI prevalence, far-reaching political and economical changes and opening of the country (Zeller et al. 1997), increasing prostitution, growing sexual tourism, poverty and local customs (blood rituals, collective circumcisions) that permit early sexual intercourse favour the evolution of HIV and other infections transmitted by similar routes.

The 18.4% prevalence rate of syphilis infection, even if overestimated, is lower than the results obtained among registered FSWs in Toliary in 1995–1996 (25.7%, n = 253; Ministère de la Santé de Madagascar 1997). It is also lower than data on FSWs from other countries such as Ethiopia where the prevalence rate of RPR positivity was 52.4% in 2000 (Aklilu et al. 2001). Our values remain distinctly higher than in other studies, for example, in Mexico the prevalence rate of RPR+ tests was 6.4% among registered FSWs in 1992 (Uribe-Salas et al. 1997). If the syphilis prevalence seems to be high, higher than in most other parts of the world, the level we found appears to be lower than in FSWs from eastern and southern Africa, where syphilis is highly endemic (Ahmed et al. 1991; Ramjee et al. 1998; WHO 2001). Our study population of registered FSWs complemented the data obtained by Behets et al. (1996) on non-official FSWs (n = 969) of Antananarivo, Taoamasina and Toliary. These unregistered FSWs were considered the most representative and perhaps most exposed to STIs, because of absence of medical and social follow-up. While expressing the limits in extrapolating data obtained in different populations, time and towns, we note that while the HIV prevalence was at the same level in these two populations, syphilis was more prevalent among unregistered FSWs (30.5%). Their syphilis prevalence rate was in better accord with data obtained by other studies on FSWs in this part of the world. The national programme aimed at controlling AIDS/STI among FSWs might have reached a certain degree of effectiveness in Toliary, as in other African countries (Ghys et al. 2001; Alary et al. 2002). However, the lack of a systematic registering procedure of the whole process in Toliary (frequency of visits, outputs of the clinical visit, reporting of the biological results) prevented the effective follow-up of these registered FSWs.

Regular condom use was unexpectedly associated with a significant risk of being RPR+ and with the self-reported markers of STIs. Even if caution has to be taken with drawing conclusions about these associations due to potential measurement bias, these paradoxical associations between the declaration of condom use and STI markers is not unusual (Mbizvo et al. 1996, 2001;McFarland et al. 1998; Behets et al. 2001). It may be the result of a methodological bias, as in cross-sectional investigations it is not possible to determine whether exposure preceded or resulted from the disease (Hennekens & Buring 1987), or it could result from an information bias, as condom use was only assessed in our study by three direct questions, although studying condom use accurately is complex (Weir et al. 1999). However, other explanations can be advanced, because a previous study reported the reluctance of the Malagasy to express their real sexual behaviours (Disaine 1993). We cannot also eliminate the possibility that only infected registered FSWs were more likely to constantly use condoms, after becoming symptomatic. Condom using could be considered a cure or a secondary prophylaxis device. Finally, even if these women constantly use condoms with their clients, they probably do not do so with their regular partners and get infected by them (according to Ravaorimalala et al. 1998; only 8.3% of the FSW population of Antananarivo declared using them with their regular companion). Our data suggest that it is also necessary to insist on condom promotion by younger registered FSWs, who seem to have difficulty in using them compared with the older ones.

In summary, this cross-sectional study highlights an unstable epidemiological situation concerning HIV infection in Toliary: on the one hand, there are no cases of HIV seropositivity among the registered FSWs of Toliary while on the other hand, due to widely distributed STIs and inadequate condom use, a strong potential exists for fast spread of an HIV epidemic. According to the Madagascar Ministry of Health, controlling STIs could lower the prospective HIV prevalence by 30% (Andriamahenina et al. 1998). The 2015 estimates of the Madagascar Ministry of Health did not deal with regional specificities, so different prospective regional prevalences could be encountered, as the 18 cultural Malagasy groups do not present the same risk factors for HIV. Because of their customs and sexual behaviour the southern populations may be at greatest risk.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References

We thank the professionals in charge of HIV/STI control in Toliary. Special thanks to all the female sex workers of the town who permitted this study to be carried out.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Laboratory exams
  7. Statistical analysis
  8. Ethical authorizations
  9. Results
  10. Study population
  11. Seroprevalence of HIV infection and syphilis
  12. Discussion
  13. Acknowledgements
  14. References
  • Ahmed HJ, Omar K, Adan SY et al. (1991) Syphilis and HIV seroconversion during a 6 month follow-up of female prostitutes in Mogadishu, Somalia. International Journal of Sexually Transmitted Disease and AIDS 2, 119123.
  • Aklilu M, Messele T, Tsegaye A et al. (2001) Factors associated with HIV-1 infection among sex workers of Addis Ababa, Ethiopia. AIDS 15, 8796.
  • Alary M, Mukenge-Tshibaka L, Bernier F et al. (2002) Decline in the prevalence of HIV and sexually transmitted diseases among female sex workers in Cotonou, Benin 1993–1999. AIDS 16, 463470.
  • Andriamahenina R, Ravelojaona B, Rarivoharilala C et al. (1998) Le SIDA à Madagascar: épidémiologie, projections, impact socio-économique, interventions. Bulletin de la Societe de Pathologie Exotique 91, 6870.
  • Behets F, Andriamahenina R, Andriamiadana J, May J et al. (1996) High syphilis and low but rising HIV seroprevalence rates in Madagascar. Lancet 347, 831.
  • Behets F, Andriamiadana J, Rasamilalao D et al. (2001) Sexually transmitted infections and associated socio-demographic and behavioural factors in women seeking primary care suggest Madagascar's vulnerability to rapid HIV spread. Tropical Medicine and International Health 6, 202211.
  • Bienzle D, MacDonald KS, Smaill FM et al. (2000) Factors contributing to the lack of human immunodeficiency virus type 1 (HIV-1) transmission in HIV-1-discordant partners. Journal of Infectious Diseases 182, 123132.
  • Chin J (1990) Surveillance du SIDA et des infections à VIH dans le cadre de la santé publique. Bulletin of the WHO 68, 699708.
  • Disaine B (1993) Connaissances, attitudes, croyances et pratiques relatives au SIDA/VIH dans le Faritany de Toliary et à Sainte-Marie. Programme national de lutte contre le SIDA, Tananarive.
  • Estébanez P, Fitch K & Nagera R (1993) HIV and female sex workers. Bulletin of the WHO 71, 397412.
  • Genin C & Coulanges P (1988) Results of a mini-survey conducted on HIV-1 seroprevalence in Madagascar. Transactions of the Royal Society of Tropical Medicine and Hygiene 82, 897.
  • Ghys PD, Diallo MO, Ettiene-Traore V et al. (2001) Effect of interventions to control sexually transmitted disease in the incidence of HIV infection in female sex workers. AIDS 15, 14211431.
  • Hennekens CH & Buring JE (1987) Descriptive study. In: Epidemiology in Medicine (ed. SLMayrent) Little Brown, Boston, pp. 108112.
  • Kanki PJ, Hamel DJ, Sankalé JL et al. (1999) Human immunodeficiency virus type 1 subtypes differ in disease progression. Journal of Infectious Diseases 179, 6873.
  • Lamptey P (2002) Reducing heterosexual transmission of HIV in poor countries. British Medical Journal 324, 207211.
  • Lopalco L, Barassi C, Pastori C et al. (2000) CCR5-reactive antibodies in seronegative partners of HIV-1-seropositive individuals down-modulate surface CCR5 in vivo and neutralize the infectivity of R5 strains of HIV-1 in vitro. Journal of Immunology 164, 34263433.
  • Mbizvo MT, Mashu A, Chipato T et al. (1996) Trends in HIV-1 and HIV-2 prevalence and risk factors in pregnant women in Harare, Zimbabwe. Central African Journal of Medicine 42, 1421.
  • Mbizvo EM, Msuya SE, Stray-Pedersen B et al. (2001) HIV seroprevalence and its associations with the other reproductive tract infections in asymptomatic women in Harare, Zimbabwe. International Journal of Sexually Transmitted Disease and AIDS 12, 524531.
  • McFarland W, Mvere D, Shamu R et al. (1998) Risk factors for HIV seropositivity among first-time blood donors in Zimbabwe. Transfusion 38, 279284.
  • Meda N (1995) La surveillance épidémiologique des MST, de l'infection par le VIH et du SIDA à Madagasacar. Rapport d'expertise WHO/GPA/AFRO. Ministère de la Santé, Tananarive.
  • Ministère de la Santé de Madagascar (1997) Données et indicateurs démographiques et socio-sanitaires à Madagascar. Mission française de coopération et d'action culturelle, Tananarive.
  • Ohshige K, Morio S, Mizushima S et al. (2000) Behavioural and serological human immunodeficiency virus risk factors among female commercial sex workers in Cambodia. International Journal of Epidemiology 29, 344354.
  • Ramjee G, Karim SS & Sturm AW (1998) Sexually transmitted infections among sex workers in KwaZulu-Natal, South Africa. Sexually Transmitted Diseases 25, 346349.
  • Ravaoarimalala C, Andriamahenina R, Ravelojaona B et al. (1998) Le SIDA à Madagascar: politique d'intervention pour maintenir la basse prévalence de l'infection par le VIH. Bulletin de la Societe de Pathologie Exotique 91, 7173.
  • Rowland-Jones SL, Nixon DF, Aldhous MC et al. (1993) HIV-specific cytotoxic T-cell activity in a HIV-exposed but uninfected infant. Lancet 341, 860861.
  • Samson M, Libert F, Doranz BJ et al. (1996) Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the CCR5 chemokine receptor gene. Nature 382, 722725.
  • UNAIDS (2000) Report on the Global HIV/AIDS Epidemic, June 2000. http://www.unaids.org.
  • Uribe-Salas F, Hernandez-Avila M, Conde-Glez C et al. (1997) Low prevalence of HIV infection and sexually transmitted disease among female commercial sex workers in Mexico City. American Journal of Public Health 87, 10121115.
  • Weir SS, Roddy RE, Zekeng L et al. (1999) Association between condom use and HIV infection: a randomised study of self reported condom use measures. Journal of Epidemiology and Community Health 53, 417422.
  • World Health Organization (2001) Global Prevalence and Incidence of Selected Curable Sexually Transmitted Infections: Overview and Estimates. WHO, Geneva. http://www.who.int/HIV_AIDS/GRSTI/index.htm.
  • Young H (1992) Syphilis: new diagnostics directions. International Journal of Sexually Transmitted Disease and AIDS 3, 391413.
  • Zeller HG, Ramamonjisoa A, Boisier P et al. (1997) HIV infection in Madagascar in 1995. AIDS 11, 401402.