GHENT UNIVERSITY Karoline FONCK - International Centre for ...

GHENT UNIVERSITY
STI PREVENTION AND CONTROL
SEXUALLY TRANSMITTED INFECTIONS IN NAIROBI, KENYA:
CLINICAL, EPIDEMIOLOGICAL AND PUBLIC HEALTH ASPECTS
Doctoral Thesis submitted to The Faculty of Medicine and Health Sciences
By
Karoline FONCK
Promotor: Prof Dr Marleen Temmerman
Dept of Obstetrics and Gynaecology
2002

C ONTENTS
List of abbreviations…………………………………………………………………………….… 2
Chapter 1. Introduction………………………………………………………………..… 4
Chapter 2. Objectives and methods………………………………………………... 18
Chapter 3. Prevalence and risk factors of STI…………………………………….. 22
3.1. Pattern of sexually transmitted diseases and risk factors among women
attending an STD referral clinic in Nairobi, Kenya………………………….... 22
3.2. Declining syphilis prevalence in pregnant women in Nairobi since 1995:
another success story in the STD field?………………………………….…… 32
Chapter 4. Health-seeking behavior and sexual behavior in Nairobi………….…… 37
4.1. Healthcare-seeking behavior and sexual behavior of patients with STDs in
Nairobi, Kenya…………………………………………………………….….. 37
4.2. Health-seeking behavior and sexual behavior among primary health care
patients in Nairobi, Kenya………………………………………………….…... 46
Chapter 5. Diagnosis and management of sexually transmitted infections…….… 56
Chapter 6. Prevention of sexually transmitted infections including HIV……. 68
6.1. Syphilis control during pregnancy: Effectiveness and sustainability of a
decentralized program……………………………………………………….…. 68
6.2. Partner notification of pregnant women infected with syphilis in Nairobi,
Kenya………………………………………………………………………….….. 73
6.3. A randomized, placebo controlled trial of monthly azithromycin prophylaxis
to prevent STIs and HIV-1 in Kenyan sex workers: study design and
baseline findings……………………………………………….….……………... 80
6.4. STIs and vaginal douching in a population of female sex workers in
Nairobi, Kenya……………………………………………………………….…... 91
Chapter 7. Conclusions…………………………………………………………….…… 99
Chapter 8. Final remarks………………………………………………………….…. 108
References……………………………………………………………………………………….… 111
Annex WHO Collaborating Center on HIV/AIDS, Nairobi, Kenya…………………... 126
Summary…………………………………………………………………………………………... 128
Résumé…………………………………………………………………………………………….. 131
Samenvatting………………………………………………………………………………….…... 135
Acknowledgements……………………………………………………………………………….. 139
Contents 1

AIDS Acquired Immune Deficiency Syndrome
BV Bacterial vaginosis
CI Confidence Intervals
CIN Cervical Intraepithelial Neoplasia
CT Chlamydia trachomatis
DSMB Drug safety and monitoring board
EC European Commission
ELISA Enzyme-linked Immunosorbent Assay
FAO Food and Agriculture Organization
FHI Family Health International
FP Family Planning
FSW Female sex workers
GoK Government of Kenya
GUD Genital Ulcer Disease
HD Haemophilus ducreyi
HGSIL High-grade SIL
HIV Human Immunodeficiency Virus
HSV Herpes simplex virus
ICRH International Center for Reproductive Health
KSh Kenyan Shilling
LED Leukocyte esterase dipstick
LGSIL Low-grade SIL
MoH Ministry of Health
M-PCR Multiplex PCR
NASCOP National Aids Control Program
NCC Nairobi City Council
NG Neisseria gonorrhoeae
NG/CT Neisseria gonorrhoeae / Chlamydia trachomatis
NGO Non-governmental Organization
NS Not significant
OR Odds Ratio
PCR Polymerase chain reaction
PHC Primary health care
PID Pelvic Inflammatory Disease
L IST OF A BBREVIATIONS
List of Abbreviations 2

PMH Pumwani Maternity Hospital
PPV Positive predictive value
RPR Rapid Plasma Reagin
RTI Reproductive tract infections
SCU STD Control Unit
SIL Squamous intraepithelial lesion
SPSS Statistical Package for Social Sciences
STC Special Treatment Clinic
STD Sexually transmitted diseases
STI Sexually transmitted infections
TP Treponema pallidum
TPHA Treponema pallidum haemagglutination assay
TV Trichomonas vaginalis
UoN University of Nairobi
VDRL Venereal Disease Research Laboratory
WHO World Health Organization
List of Abbreviations 3

1.1. Sexually Transmitted Infections
C HAPTER 1
INTRODUCTION
Sexually transmitted infections (STIs) have, in all likelihood, always been present. The recent
emergence of HIV/AIDS has however added new relevance to STIs and has also reinforced
the tremendous cultural burden of diseases associated with sexual contact.
Sexually Transmitted Infections are a group of communicable diseases that are
predominantly transmitted through sexual contact. In many countries STIs have become a
major public health problem on account of their frequency, their impact on maternal and
infant health, their personal and social consequences as well as their cost in terms of health
expenditure and lost productivity. Women, and neonates in particular, carry the major burden
of morbidity and mortality due to STI and or their ensuing complications. The World Bank has
estimated that, in 1990, STIs, excluding HIV infection, rank second, after all maternal causes,
as a cause for healthy life lost in women aged 15 to 44.
Africa carries the largest part of the HIV/AIDS pandemic burden. According to UNAIDS
estimates, over 23 million people were living with HIV in Africa at the end of 1999. The
predominant mode of transmission in Africa has always been heterosexual. The morbimortality
owing to HIV/AIDS has major impacts. The impact on the health sector is
overwhelming. Furthermore, AIDS is systematically cutting down life expectancy in the
countries where the disease is most common. Patterns of HIV infection however show a
great deal of variation with some countries beginning to show stabilization while other
countries experience rapid growth in infection rates. The stabilization is a natural part of the
epidemic process and occurs as a natural result of a dynamic balance between new
infections and deaths.
1.2. Transmission dynamics of STIs
STIs are transmitted through a complex interplay among the natural infectiousness of the
pathogen, the duration of infectivity of the human host, and the rate of sexual interaction
between people.
Chapter 1 4

Several models have been developed for STI epidemiology. The essential concept in these
models is the reproductive rate (R0) of the infection. R0 is defined as the number of
secondary infections that a spreader produces in a fully susceptible population. For STI
transmission, R0 is determined by three variables. These are the average rate at which new
sex partners are acquired per unit of time (c), the average probability that infection is
transmitted from one person to a susceptible individual per sexual contact (ß), and the
average duration of infectiousness of an infected person (D). In the simplest possible case,
where a single risk group mixes homogeneously (chooses sex partners at random), the
transmission dynamics are approximated by the formula:
R0 = ßcD
What is meant exactly by average is not straightforward, since epidemiological studies reveal
much variability between and within specific populations. Many heterogeneities influence the
magnitude of β (sex acts per unit of time, type of sex act) and measures of this value via
studies must be interpreted with caution. Recorded estimates of the magnitude of β range
from a very high likelihood of transmission for infections such as gonorrhoea and syphilis and
to a low probability for infections such as HIV and chlamydia. The typical duration and
degree of infectiousness (D) is also subject to variation and duration often varies widely
between individuals. The duration of infectiousness can be influenced by interventions such
as chemotherapy. Treatment therefore reduced the value of R0.
The principal determinant of STI spread in a population is the rate at which new sexual
relationships occur in that population, rather than the number of sexual exposures per new
partnership. The more sexual partners one has, the greater the likelihood of encountering
high frequency transmitters of STI pathogens. Rates of sexual partner change vary widely
within and between societies, and are associated with a range of demographic and socioeconomic
factors. Invariably, the reported rate of change of sex partners for men exceeds
that for women. Surveys reveal much heterogeneity with most people reporting few partners
and a few reporting many. That core group of highly sexually active individuals however is a
major influence on the transmission pattern, as they will assure that R0 exceeds unity in
value (Yorke 1978, Anderson 1987, Anderson 1988).
Individuals with two or more simultaneous (i.e. concurrent) partners can play a central role in
the spread of infection, as the individual can acquire infection from one partner and pass it on
to the another without gaining any new partners. In both developed and developing
INTRODUCTION 5

countries, the rate of reporting concurrent sex partners is often high. In both cases, the
frequency is higher for males than for females. The existence of a high prevalence of
concurrent partners in the population is of particular importance in ensuring the persistence
of infection in low-risk groups, as this may assure “bridging” of the populations to spread
infection from high to low prevalence subgroups within a population.
The components that contribute to success of the transmission will influence the
epidemiological pattern in different ways. Some of the key elements are: variation in
infectiousness to sexual partners over the incubation period of an STI in the index case; the
structure of the mixing matrix i.e. the sexual partner network; presence of co-factors that
enhance transmission.
Understanding the transmission dynamics in a population provides insight into how different
interventions are likely to influence transmission success and consequently the prevalence
and incidence of infection in a given population.
1.3. Epidemiology of STI and HIV
It is estimated that the overall number of new cases of the major four STIs for men and
women aged 15 to 49 totaled 340 million in 1999, with 12 million cases of syphilis, 62 million
of gonorrhea, 92 million of chlamydial infections and 174 million of trichomoniasis (WHO
2001). These estimates suggest that 90 percent of these STIs occur in developing countries.
The highest rate of new cases per 1000 population occurred in sub Saharan Africa. The
prevalence rates of some STIs in countries in sub-Saharan Africa are represented in Table 1.
Table 1: Prevalence of N. gonorrhoeae (NG), C. trachomatis (CT) and positive syphilis serology
(VDRL or RPR) among pregnant women in selected developing countries in sub-Saharan Africa.
Country Syphilis (%) NG (%) CT (%)
Kenya (Temmerman 1992)
5.3
6.5 (Jenniskens 1995)
Botswana (Pedersen 1990)
Rwanda (Leroy 1995)
16.8 14
HIV +ve
6.3
2.4
HIV –ve
3.7
7.1 9.0 (Temmerman 1992)
Tanzania (Mayaud 1995) 10.1 2.1 6.6
Ethiopia 13.7 (Azeze 1995) 29 (Duncan 1992)
Congo (Yala 1991) 9 26.8
Lesotho (Fehler 1995) 12 5 14
South Africa (Rotchford 2000) 9 4.7 (Dietrich 1992)
Congo (Blanchard 1999) 6.7 3.1 6.2
INTRODUCTION 6
3.4
5.5

In developing countries, the STI epidemiology is influenced by a number of socio-cultural and
economic conditions, making STIs one of the main public health priorities (Van Dam 1999).
The incidence and prevalence of STIs are generally high both in urban and rural populations,
although there is considerable variation. Rates of complications and long-term sequelae are
high as well because diagnosis and effective treatment are often delayed. Substantial gender
inequalities exist which render women more vulnerable to STIs and reduce women’s access
to information and health care services. STI care is provided by a large variety of health care
providers, many of who are poorly trained in STI case management and the quality of care is
questionable. Healthcare-seeking behavior is often inadequate owing to frequently
asymptomatic nature of STI in women, the generally low awareness of genital health and the
stigma associated with genital symptoms. Public resources are limited and health care
budgets are low.
1.4. STI and HIV interaction
The interaction between STI and HIV is complex. Both HIV and the traditional STIs share a
common mode of transmission through sexual contact. They share the same behavioral
risks. The reciprocal relation or “epidemiological synergy” between HIV infection and other
STIs has been described (Wasserheit 92, Fleming 1999). First, HIV induces alterations in the
natural history, diagnosis or response to therapy of other STIs, although more so for
ulcerative STIs. Second, an array of studies has shown that both ulcerative and nonulcerative
STIs increase the risk of HIV transmission and acquisition, although non-ulcerative
STIs increase risk primarily for the receptive partner. Still, since the prevalence and
incidence of non-ulcerative STIs is much higher than genital ulcer disease (GUD) in many
populations, their impact on HIV transmission may be more important on a population level
(Piot & Holmes 1984, Laga 1993, Clottey & Dallabetta 1993).
As observational studies are inherently subject to a number of potential limitations and they
cannot directly measure the effect of STI control on HIV incidence in the larger population,
community level randomized controlled intervention trials have been conducted. In Mwanza,
Tanzania, improved continuous management of symptomatic bacterial STI at the primary
health care level was introduced. After 24 months, this intervention resulted in a 38%
reduction in HIV incidence in the intervention arm. Significant reductions were documented in
the prevalence of serologically diagnosed syphilis and of symptomatic urethritis in the
previous year. These results were not associated with changes in sexual behavior or condom
use (Grosskurth 1995, Mayaud 1997). A subsequent HIV-intervention trial in the Rakai
INTRODUCTION 7

district of Uganda randomized villages to mass treatment of STIs every 10 months versus
placebo. After three mass treatment rounds that spanned 20 months, the study showed no
effect on HIV incidence. Significant reductions were observed in the prevalence of
serologically diagnosed syphilis and trichomoniasis and in the incidence of trichomoniasis but
not in syphilis (Wawer 1999).
The differences in outcome in the two studies are intriguing but are not invalidating the
effectiveness of STI control (Fleming 1999). At least four explanations contributed to the
divergent results. First, continuous access to improved STI treatment services would be
more effective than intermittent mass treatment. Second, symptomatic STIs may be more
important than asymptomatic infections in facilitating HIV transmission, as they often reflect
recently acquired or incident infections and therefore, recent risky sexual behavior. Third,
STI may play a greater role in HIV transmission in earlier than in later phases of an HIV
epidemic. As HIV epidemics mature, exposure becomes increasingly independent of
cofactors as STI. The Rakai district was experiencing a far more advanced HIV epidemic
than the Mwanza region with HIV baseline prevalence rates of 16% and 4% respectively.
Finally, STI incidence, prevalence and etiological spectrum are critical determinants of the
impact of any STI treatment intervention on HIV transmission. Available data suggest that
while syphilis prevalence rates were comparable, gonorrhea and chlamydia prevalence may
have been slightly higher in Mwanza, and the proportion of genital ulcers due to HSV-2 was
probably substantially higher in Rakai.
The extensive observational and intervention trial data leave little doubt that other STIs
facilitate HIV transmission. Hence, interventions, which lead to the reduction of the levels and
duration of these STI, should have an impact on the HIV epidemic. Therefore, the question is
no longer whether STI detection and treatment should be an essential component of HIV
prevention programs but rather how this component should be implemented to have maximal
impact. WHO recognized the need for a common approach in AIDS and STI prevention and
control programs as 1) the predominant mode of transmission of HIV and STI is sexual; 2)
many measures of prevention are the same, as the target audiences for these interventions;
3) clinical services for STIs are important points of contact with persons at high risk; 4) early
diagnosis and treatment of STIs are an important strategy for prevention of HIV transmission;
5) trends in STI incidence and prevalence are easier to monitor than HIV seroprevalence or
incidence and can be useful early indicators of changes in sexual behavior (WHO 1997,
1999).
INTRODUCTION 8

1.5. STI prevention and control
The main objectives of STI control are: 1) to interrupt the transmission of STI; 2) to prevent
the development of complications and sequelae; 3) to reduce the incidence of HIV infection.
These objectives can be achieved by programs through primary and secondary prevention.
Primary prevention aims at reducing the incidence of disease through prevention of
acquisition of infection and disease. This can be done through safer sexual behavior and
condom use. Secondary prevention is directed at reducing prevalence by shortening the
duration of disease thus minimizing the probability of complications or sequelae, through the
provision of treatment and care for infected and affected persons.
Growing emphasis has been placed on behavioral interventions to reduce exposure, in view
of the growing importance of incurable viral STIs including HIV infection. A comprehensive
priority public health package for STI control in developing countries should include 1) health
promotion to reduce risk of exposure to infection and adoption of “safe sex” practices,
including the use of condoms and the maintenance of these safer behaviors; 2) promote
early STI healthcare-seeking behavior; 3) adequate management of patients with STIs and
their partners; 4) intensified interventions in population groups with highest rates of risk
behavior. For the implementation of this package it is essential that a supportive national
policy be adopted that endorses and elaborates these components (WHO 1999).
Both for STI and HIV prevention and control, it is essential that health education and
promotion efforts be intensified and sustained to achieve an urgently needed change in risktaking
behavior, to maintain safe behaviors and to develop an environment that enables
people to adopt and sustain safe behavior. Many societies however are reluctant to openly
address issues involving sex and sexuality and to recognize the realities of sometimes
widespread existence of pre- and extramarital sexual intercourse.
Furthermore, the effect of poor healthcare-seeking behavior coupled to ineffective STI
services result in lack of a major impact in view of the large number of people with infection
who fail to obtain treatment. (WHO 1997, 1999). As described in the Operational model of
effectiveness of STI services, also called the Piot-Fransen model (Figure 1), the success of
an STI program should be measured in terms of the proportion of cases of STI occurring in
the population that is successfully treated, considering that STI infected individuals will have
to overcome a series of hurdles before they can be considered cured. The model shows
“leakage” at each step, determining what proportion of patients is lost from treatment at
successive points in the process from infection to possible cure.
INTRODUCTION 9

By identifying where significant “leakages” take place in a given setting, priorities for
improved STI service delivery can be set.
Figure 1. Operational model of effectiveness of STI services (Piot-Fransen model)
Number of persons
infected
Number aware of
infection
Number Seeking
appropiate health care
Number Diagnosed with
STD
Number Receiving
treatment
Number cured
Number of infected
people not cured
The work presented here focuses on the epidemiology of STI including HIV in women, men
and children in Kenya, particularly Nairobi, and the impact as well as the effectiveness of
some control interventions that are in place. These findings may contribute to improved
strategies for both the prevention and control of STI in Kenya.
1.6. Study site
The republic of Kenya is located in East Africa and is
bordered by Somalia, Ethiopia, Sudan, Uganda,
Tanzania and the Indian Ocean. Its surface is
582,446 km 2 (Figure 2). The population is originating
from different ethnic groups, mainly of Bantu origin.
Figure 2: Map of Kenya (source: CIA)
INTRODUCTION 10

In the 1980s Kenya had the fastest population growth rate in the world and increased by a
third in a decade to 1999 to reach over 30 million (FAO/World Bank, 1999) (Figure 3). Data
indicate that the birth rate has plummeted from 8.1 live births per woman in 1978, to 4.6 in
1998 and to 3.7 in 2000 (US census bureau). These declines can partially be attributed to
increased use of contraceptives, but the HIV/AIDS epidemic may also play a role. Overall life
expectancy was 59 years (1989-94). The crude birth rate is 29 per 1000 population, the
crude death rate 14 per 1000 population and the infant mortality rate 69 per 1000 live births.
Mortality rates are increasing due to aids-related deaths. The capital Nairobi, with over 2
million inhabitants, is expanding rapidly due to internal migration from rural areas. A large,
but uncounted, proportion of the urban population lives in shantytowns.
Figure 3: Population pyramid, Kenya 1998
Sexually Transmitted Infections (STI) have been recognized as a major public health
problem in Kenya. STI are among the most common complaints of adults at outpatient
clinics, representing up to 10% of the caseload at many health facilities. The overall
prevalence of these infections in the general population in Nairobi is unknown, but among
pregnant women the seroprevalence rate of syphilis is about 3 to 5%, and the prevalence of
gonorrhea is about 2 to 8%, chlamydia 8.8%, trichomonas 19.9% (Laga 1986, Temmerman
1992, Thomas 1996, Gichangi 1997). In a group of family planning attendees in Nairobi,
prevalence for gonorrhoea was 3.2%, syphilis serology 1.9% and trichomonas 5.2% (Daly
1994). In a study among male transport workers in Mombasa, eastern Kenya, the prevalence
of gonorrhoea was 3.4%, chlamydia 3.6% and trichomonas 6% (Jackson 1997). A recent
population based multi-center study revealed that in Kisumu, western Kenya, prevalence
rates for men were 0% for gonorrhoea, 2.6% for chlamydia, 3.4% for syphilis, 34.6% for
HSV-2 and 19.8% for HIV. For women, those prevalence rates were respectively 0.9%,
INTRODUCTION 11

4.5%, 3.9%, 67.7% and 30.1% while trichomoniasis was prevalent in 29.3%. The prevalence
for HIV infection among CSWs in that community was 74.7% (Buve 2001).
The consequences of untreated STIs are also a major cause of concern. Gonorrhea and
chlamydia are major causes of pelvic inflammatory disease (PID) and infertility among
women, as well as of urethral stricture among men. In Africa, 64% of women in infertile
couples have tubal abnormalities and/or pelvic adhesions (Cates 1985). In Kenya, PID is an
important cause of tubal infertility (Cohen 2000). Untreated maternal syphilis results in
neonatal morbidity and mortality in 60% of the cases (Harman 1917). A related problem in
southern and eastern African countries is cancer of the cervix, the most frequently detected
cancer among African women. Cervical dysplasia rates of 2% to 12% have been reported in
selected populations in Kenya (Maggwa 1993, Engels 1992, Temmerman 1999).
HIV/AIDS has taken on devastating proportions in Kenya. Since the first indigenous case of
AIDS was diagnosed in 1984, the epidemic has known a rapid rise. Transmission occurs
primarily through heterosexual contact and through mother-to-child transmission. The overall
HIV-1 seroprevalence rate was estimated to be around 14% in 1999 (UNAIDS 2000). The
same report appraises the HIV prevalence among pregnant women between 13 and 35%. It
is estimated that over 2 million adults and children in Kenya are living with HIV/AIDS. It was
estimated that in 1995 about 20% of the MoH budget was consumed by HIV/AIDS (FHI
1996). However, it has not been until 1999 that AIDS was declared a “national urgency” by
the Kenyan president.
The Government of Kenya (GoK) has made commitments to fight the AIDS epidemic and to
control STI in general. In 1987, the GoK established a National AIDS Control Program
(NASCOP) within the MoH. In addition to government support, considerable funding has
been forthcoming from several bilateral and multilateral donor agencies.
A Sexually Transmitted Diseases Control Unit (SCU) was formed within the MoH in 1988,
with funding provided by the Government of Kenya and the Government of Belgium. The
SCU was charged with formulating, initiating and coordinating STI prevention and treatment
at a national level. Later the SCU was merged with the NASCOP to become a single unit.
The Belgian assistance to the NASCOP continued until 2000.
This work took place at the WHO Collaborating Center for Research and Training in
STD/AIDS, at the Department of Medical Microbiology, University of Nairobi, between 1996
and 2000. At the Department of Medical Microbiology of the University of Nairobi, an almost
INTRODUCTION 12

state-of-the-art laboratory for the detection of STI and HIV is available. The clinical work is
performed mainly at following sites (Figure 4):
• The Nairobi Special Treatment Clinic (STC), a down-town City Council referral STI
clinic receiving 150 to 200 patients per day;
• Pumwani Maternity Hospital, a large City Council referral maternity hospital with an
average of 75 deliveries per day;
• The Pumwani prostitute clinic, a special out-patient clinic for commercial sex workers
in a slum area where commercial sex workers live and work;
• The Langata City Council health center in which premises an out-patient clinic for a
cohort of commercial sex workers from the surrounding largest Nairobi slum area has
been established;
• The ten City Council clinics all around the city of Nairobi, with decentralized STI case
management, where first line STI treatment is provided including RPR screening for
syphilis in pregnancy;
Figure 4: NCC health facilities in Nairobi
Source: HassConsult real estate
1.7. Study populations
1 Ngong Road
2 Riruta
3 Langata
4 Westlands
5 STC
6 Pumwani Maternity
7 Eastleigh
8 Umoja
9 Kariobangi
10 Baba Dogo
11 Dandora
The study populations consist of various target groups such as commercial sex workers
(core group), male and female STI clinic attendees (possible “bridging” group), and male and
INTRODUCTION 13

female primary health care or antenatal clinic attendees (general population) as summarized
in Table 2.
Table 2: Summary of different population groups studied.
Place of recruitment Sex Number Some characteristics
Study 1 STC Female STI clinic attendees 520 33 % pregnant
Study 2 PHC Pregnant women 81,311 5.2% RPR positive
Study 3 STC Male and female STI clinic attendees 471 50% male
Study 4 PHC Male and female PHC attendees 555 44% male
Study 5 PHC + STC Women with vaginal discharge 621 53 % pregnant
Study 6 PHC Pregnant women 27,377 3.4% RPR positive
Study 7 PMH Post-delivery syphilis sero-positive women 357
Study 8 PHC Commercial sex workers 318 5% pregnant
Study 9 PHC Commercial sex workers 543
A total of 54 health centers, controlled by the Public Health Department of the Nairobi City
Council, care for the largest part of the low-income population of Nairobi. Mission and nongovernmental
organizations (NGO) health facilities also exist in Nairobi, especially in the
slum areas. It is not known what proportion of the population is covered by their services.
Recently, Kenya, and in particular Nairobi, has seen a boom in private health facilities, going
from private hospitals to nurse-run medical cabinets. Very often, these facilities are staffed by
personnel employed by the MoH or the NCC who try to increase their rather modest public
service salary by accumulation of employment. Although no data is available on the
utilization rate of these centers, it seems that they are quite popular. This might be one of the
reasons why the STI clinic catered for about 150 - 200 patients per day at the time of the
study, reduced from about 500 daily a couple of years earlier.
1.8. STI management in Nairobi
STI case management has been decentralized in Nairobi to ten primary health care (PHC)
clinics. Those ten clinics have trained staff and are provided with the necessary first line STI
drugs. Diagnostic algorithms are used at this level for diagnosis and treatment of STIs.
These algorithms provide also for one alternative treatment after a first treatment failure. If
still unsuccessful, the patient is referred to the STI clinic in downtown Nairobi, the STC
(Special Treatment Center). PHC clinics without possibilities of STI management may refer
the patient either to a nearby PHC with STI management or directly to the STI clinic (STC), a
decision mostly based on distance.
INTRODUCTION 14

The STI referral clinic, STC, is located in downtown Nairobi and offers STI syndromic
management, laboratory services, health education and HIV counseling and testing. An
average of 200 STI patients per day was cared for, the majority being male. In November
1996, a flat fee of 100 KSh was introduced, resulting in a drop in the number of patients
(about 150 per day), which picked up slowly by 1998, at which stage the number of women
bypassed the male patients attending STC. The majority of patients are self-referred (65% in
1998). Of all new cases seen, 10% to 12% are contacts of index patients at the same clinic.
One of the priorities of the Kenyan authorities was to control syphilis during pregnancy and a
centralized program was set up to that effect. After an evaluation of that program revealed
some problems, syphilis screening has equally been decentralized in Nairobi. The same ten
primary health care clinics have been equipped to perform a rapid plasma reagin (RPR) test
on site. Pregnant women attending other PHC are referred to these 10 clinics for syphilis
testing.
INTRODUCTION 15

The research data have been published in the following papers in international peerreviewed
journals:
1. Fonck K, Kidula N, Kirui P, Ndinya-Achola J, Bwayo J, Claeys P, Temmerman M. Pattern
of sexually transmitted infections among women attending a major STI referral clinic in
Nairobi, Kenya. Sex Transm Dis 2000; 27:417-23. (Chapter 3)
The study was undertaken to 1) determine the etiology of STI among women attending
the STD referral clinic, 2) determine the prevalence of CIN in this population, and 3) study
risk markers for STIs and CIN.
2. Temmerman M, Fonck K, Bashir F, Inion I, Ndinya-Achola J, Bwayo J, Kirui P, Claeys P,
Fransen L, Declining syphilis prevalence in pregnant women in Nairobi since 1995:
another success story in the STD field? Int J STD AIDS 1999; 10:405-8. (Chapter 3)
This paper describes the trend in syphilis seroprevalence in pregnant women between
1994 and 1997.
3. Fonck K, Mwai C, Rakwar J, Kirui P, Ndinya-Achola J, Bwayo J, Temmerman M.
Healthcare-seeking behavior and sexual behavior of patients with sexually transmitted
diseases in Nairobi, Kenya. Sex Transm Dis 2001; 28:367-371. (Chapter 4)
The objective of this study was to identify health-seeking behavioral factors and health
service related issues that can be used to improve STI services, and to develop
education campaigns for the general public about seeking timely and correct STI
treatment.
4. Fonck K, Mwai C, Bwayo J, Achola J, Temmerman M. Health-seeking and sexual
behaviors among health care patients in Nairobi, Kenya. Sex Transm Dis 2002; 29:106-
111. (Chapter 4)
The objective of the study was similar to the previous study, but different results were
expected as a different population was targeted.
5. Fonck K, Kidula N, Jaoko W, Estambale B, Claeys P, Kirui P, Bwayo J, Temmerman M.
Validity of the vaginal discharge algorithm among pregnant and non-pregnant women in
Nairobi, Kenya. Sex Transm Inf 2000; 76:33-38. (Chapter 5)
This study was undertaken to validate the current Kenyan clinical algorithm for vaginal
discharge. We validated additional risk scores as well as inclusion of signs and
symptoms among pregnant and non-pregnant women.
INTRODUCTION 16

6. Fonck K, Claeys P, Bashir F, Bwayo J, Fransen L, Temmerman M. Syphilis control in
pregnancy: effectiveness and sustainability of a decentralized program, Am J Pub Health
2001; 91:705-7. (Chapter 6)
This paper provides an analysis of the existing syphilis program in Nairobi.
7. Gichangi P, Fonck K, Sekande-Kigondu C, Ndinya-Achola J, Bwayo J, Kiragu D, Claeys
P, Temmerman M. Partner notification of pregnant women infected with syphilis in
Nairobi, Kenya. Int J STD & AIDS 2000; 11:257-61. (Chapter 6)
In this paper, we report the determinants and effect of partner notification and treatment
in pregnant women.
8. Fonck K, Kaul R, Kimani J, Keli F, MacDonald KS, Ronald AR, Plummer FA, Kirui P,
Ngugi EN, Moses S, Temmerman M. A randomized, placebo-controlled trial of
azithromycin prophylaxis to prevent STI and HIV-1 in Kenyan sex workers: study design
and baseline findings. Int J STD AIDS 2000; 11:804-11. (Chapter 6)
The overall objectives of this study were to 1) evaluate the impact of regular routine
azithromycin chemoprophylaxis on STI and HIV-1 incidence in this population; 2)
examine potential adverse effects of monthly azithromycin administration, including
effects on antibiotic resistance patterns in bystander flora.
9. Fonck K, Kaul R, Keli F, Bwayo J, Ngugi E, Moses S, Temmerman M. Sexually
transmitted infections and vaginal douching in a population of female sex workers in
Nairobi, Kenya. Sex Transm Inf 2001; 77:271-276. (Chapter 6)
This study was undertaken to assess the association between vaginal douching and
sexually transmitted infections among a group of female sex workers.
INTRODUCTION 17

2.1. OBJECTIVES
2.1.1. General objectives
C HAPTER 2
OBJECTIVES AND METHODS
The overall aim of this work was to assess the epidemiology of STI including HIV-1 in groups
at various risk for HIV/STI and to attempt to document the quality, impact and effectiveness
of existing STI control interventions in Nairobi. This information in turn should help national
authorities strengthen their STI prevention and control strategies.
2.1.2. Specific objectives
1. To study the prevalence rates of sexually transmitted pathogens including HIV in
different populations in Nairobi, Kenya;
2. To examine the effects of some of the risk factors for STI transmission;
3. To assess the health-seeking behavior among the population in Nairobi in relation to
STI;
4. To define the effectiveness of the existing treatment algorithm in use in Kenya;
5. To assess the impact and the cost effectiveness of the existing syphilis screening
program for pregnant women in Nairobi;
2.2. METHODS
2.2.1. Study design
The studies took place within the framework of the EC funded AIDS/STD Control Project in
Nairobi and Mombasa. As part of this project, operational research aimed at improving
existing prevention and control strategies, was performed.
We used observational studies including prospective cohort and cross-sectional designs.
Routinely collected data from clinics and labs were also used. The methods used and the
population examined are described in detail in each chapter.
Chapter 2 18

All studies were approved by the ethical committee of the University of Nairobi. No additional
ethical approval from another institution was sought, except for the randomized placebocontrolled
trial. This trial was executed in collaboration with the University of Manitoba,
Winnipeg, Canada, and received additional approval from the ethical committee of that
University. A drug and safety monitoring board (DSMB) was established to monitor that study
closely. Informed consent was obtained from all study participants.
2.2.2. Data collection
a. Staff
The core staff involved in the research studies consisted of two local physicians, a social
scientist, a nurse, a counselor, a data entry person, one administrator, one clerical staff and
one driver. Whenever necessary, extra staff was hired for temporary assignments. These
persons were mainly recruited among the existing University of Nairobi staff or from the
Nairobi City Council.
Data collection was performed by trained staff using hospital and clinic record cards as well
as standard questionnaires specifically designed for these clinical epidemiological studies.
b. Quality of the data
Substantial consideration was given to the method of data collection. Many of the standard
interview forms also catered for behavioral and sexual data. To avoid incorrect answers on
these sensitive issues, experts in communication have been involved in the preparation of
the questionnaires. Before the start of the studies, the questionnaires were field tested and
adapted. Research assistants were trained in the use of the questionnaires. However, no
quality control for questionnaire administration was done. It can be presumed though that if
systematic errors occurred, they should be similar in HIV positive and negative persons, as in
almost all cases the patients were not aware of their serostatus.
A copy of the study proposals, standard questionnaires and data entry forms are available
upon request.
OBJECTIVES AND METHODS 19

2.2.3. Laboratory procedures
The microbiological and serological laboratory methods are described in detail in the different
chapters. The samples were collected by the research staff and transported to the research
laboratory at the Department of Medical Microbiology, University of Nairobi. Here thoroughly
trained personnel performed the tests in excellent conditions and with up-to-date equipment.
The laboratory was a research lab only and did not serve other hospitals or clinics. However,
because of cost-recovery this has changed in the last couple of years of the study. The lab is
funded by Canadian, American, European Union, Belgian and Kenyan research grants. All
senior laboratory staff has been trained in overseas laboratories of the collaborating
institutions.
Some tests were also performed at the laboratory of the STI referral clinic in Nairobi. This lab
had been considerably upgraded by the project and qualified staff had been hired. This lab
performed the quality control for the peripheral health centers for instance, all positive RPR
tests as well as 10% of the negative samples were retested at STC. The samples from the
STI clinic were in turn controlled by the lab at the University of Nairobi.
Some of the samples were also sent to the laboratories of the collaborating universities
overseas, either to perform the tests that could not be done on site, or for quality control on a
random selection of samples. Internal quality control was performed on a regular basis.
2.2.4. Data analysis
Along with the questionnaire, a computerized database structure was created for data entry.
Data was entered at the project offices either at the STI clinic or at the University of Nairobi.
Several computers were available to that effect. The data entry was supervised by the
respective research staff and under the overall supervision of the project manager.
The data from the earliest studies were entered in Epi-info (version 6.0) and afterwards
exported to SPSS for analysis. Later on, most data were entered and analyzed in SPSS for
Windows version 8.0 (SPSS Inc. Chicago). For one study the data were first entered in
ACCESS Microsoft and later exported into SPSS. Data were entered continuously
throughout the studies. Interim analysis was performed regularly to monitor the study and
discussed in meetings with the research staff.
OBJECTIVES AND METHODS 20

Once data entry was completed, they were checked through comparison with the
questionnaires. Preliminary data analysis permitted to detect errors in data entry. For the
larger studies, double data entry was performed. After data cleaning, analysis took place in
Nairobi but also in the overseas collaborating universities. Results were discussed with
scientific staff including a statistician, epidemiologists, behavioral scientists, fellow
researchers etc.
In univariate analysis, the Odds Ratios were used for the measurement of association of
proportions. Comparisons were made using Pearson’s χ 2 and Fisher’s exact tests. T test was
used for comparison of means. Logistic regression was used to take into account risk factors
and presence of other STIs for the risk of HIV infection.
2.2.5. Data dissemination
As the main aim of the research studies was to assess the quality and effectiveness of
existing STI prevention and control interventions, in order to be able to improve on them, it
has always been a priority to disseminate the results of the studies to the interested parties.
The results were discussed at the monthly steering committee where the National Aids
Control Program (NASCOP, MoH) and the European Commission were represented besides
the main partners, the University of Nairobi and the Nairobi City Council. Copies of finalized
papers were sent to interested parties including the NASCOP.
2.2.6. Conclusion
Consideration was given to issues such as quality of data collection and data entry, training
of research staff, quality control of laboratory results and data analysis. Nevertheless,
because of the methodological difficulties specific to this kind of research, some drawbacks
were unavoidable (logistic and administrative problems, loss to follow-up) and will be
discussed in detail in each chapter as well as in the general conclusion.
OBJECTIVES AND METHODS 21

C HAPTER 3
PREVALENCE AND RISK FACTORS OF STI
3.1. Pattern of Sexually Transmitted Diseases and Risk Factors
Among Women Attending an STD Referral Clinic in Nairobi, Kenya
Summary
Published in Sexually Transmitted Diseases 2000; 7:417-23, by
Fonck K, Kidula N, Kirui P, Ndinya-Achola J, Bwayo J, Claeys P
and Temmerman M.
In Kenya, sexually transmitted disease (STD) clinics care for large numbers of patients with
STD-related signs and symptoms. Yet, the etiologic fraction of the different STD pathogens
remains to be determined, particularly in women. The aim of the study was to determine the
prevalence of STDs and of cervical dysplasia and their risk markers among women attending
the STD clinic in Nairobi. A cross-section of women were interviewed and examined;
samples were taken. The mean age of 520 women was 26 years, 54 % had a stable
relationship, 38% were pregnant, 47% had ever used condoms (1% as a method of
contraception), 11% reported multiple partners in the previous 3 months, and 32% had a
history of STDs. The prevalence of STDs was 29% for HIV type 1, 35% for candidiasis, 25%
for trichomoniasis, 16% for bacterial vaginosis, 6% for gonorrhea, 4% for chlamydia, 6% for a
positive syphilis serology, 6% for genital warts, 12% for genital ulcers, and 13 % for cervical
dysplasia . Factors related to sexual behavior, especially the number of sex partners, were
associated with several STDs. Gonorrhea, bacterial vaginosis, cervical dysplasia, and genital
warts or ulcers were independently associated with HIV infection. Partners of circumcised
men had less-prevalent HIV infection. Most women reported low-risk sexual behavior and
were likely to be infected by their regular partner. HIV and STD prevention campaigns will not
have a significant impact if the transmission between partners is not addressed.
Chapter 3 22

Introduction
It is now well documented that the presence of ulcerative and non-ulcerative sexually
transmitted diseases (STDs) facilitates the transmission and acquisition of HIV (Wasserheit
1992, Sewankambo 1997, Schacker 1998, Plummer 1991, Kreiss 1994, Moss 1995, Laga
1993). Therefore, STD control has been recognized as a key strategy to reduce HIV infection
(Laga 1994, Grosskurth 1995). Furthermore, STD and their ensuing complications are a
major burden of morbidity and mortality, particularly among women and neonates in
developing countries (Piot 1990, Ryder 1991, Temmerman 1992, Temmerman 1990). A
related problem in southern and eastern African countries is cancer of the cervix, the most
frequently detected cancer among African women (Mati 1984). Cervical dysplasia rates of
2% to 12% have been reported in selected populations (Maggwa 1993, Engels 1992).
In Kenya, STD treatment using syndromic management has been introduced in different
primary health care centers. Health care providers were trained and first-line drug kits were
distributed in the health sector. Patients requiring second-line treatment or patients with
complicated reports of symptoms had to be referred to one of the STD referral clinics. The
STD clinic in downtown Nairobi is well equipped with supplies, drugs, and a trained staff, and
is well known among the local population. This study was undertaken to identify the burden
and pattern of disease in the STD referral clinic in Nairobi. The specific objectives were to
(1) determine the prevalence of sexually transmitted infections among women attending the
main STD referral clinic in Nairobi, (2) determine the prevalence of cervical intraepithelial
neoplasia (CIN) in this population, and (3) study risk markers for STD, HIV, and CIN in this
population.
Methods
Population Examined and Clinical Assessment
The target study group included women attending the Special Treatment Clinic, the main
STD referral clinic in Nairobi, Kenya. The Special Treatment Clinic is well known in Nairobi,
particularly among persons with a lower socio-economic background. Many self-referred
patients also frequent the clinic.
The average attendance in the clinic is approximately 150 patients per day, half of whom are
women. After registration, patients are referred to the clinical officers in one of the
consultation rooms. Study enrollment took place from June 1996 to April 1997.
PREVALENCE AND RISK FACTORS OF STI 23

After registration by the clinic clerks, the patients were taken inside the clinic. The choice of
the physician who would see the patient depended solely on the number of patients still
queuing in front of the clinician's room. The selection of the patients was in that sense not
entirely random, but the best alternative given the circumstances. The women who were
willing to participate in the study were examined by the research physician. Informed consent
was obtained from all participants. A nurse was responsible for administering a standardized
questionnaire to each woman, and also offered pretest HIV counseling. Data collection
included socio-demographic data, previous reproductive medical history, sexual behavior,
and the presence of genital symptoms. The study physician performed a full gynecologic
examination. Vaginal swabs were taken for wet mount, potassium hydroxide testing, and pH
testing. Cervical samples were obtained for Gram staining and detection of Neisseria
gonorrhoeae and Chlamydia trachomatis. A Papanicolaou smear was taken, and venous
blood was drawn for rapid plasma reagin (RPR) and HIV testing. The patients were treated
according to the national guidelines, following the syndromic approach. During a follow-up
visit the treatment was adapted, when necessary, according to the laboratory findings.
Laboratory Techniques
Saline wet mounts were examined for the presence of motile Trichomonas vaginalis, of yeast
cells indicative of Candida albicans, and of clue cells indicative of bacterial vaginosis. The
analysis of other samples was performed at the laboratory of the Department of Medical
Microbiology (University of Nairobi). Bacterial vaginosis was diagnosed, based on the
following symptoms: vaginal fluid pH of more than 4.5, release of a fishy amine odor from
vaginal fluid mixed with 10% potassium hydroxide, and presence of clue cells on wet mount.
Swabs for N. gonorrhoeae isolation were inoculated onto a Thayer-Martin medium and
incubated in a candle extinction jar at 36 °C for 24 hours to 48 hours. Swabs for chlamydia
were tested with enzyme-linked immunoassay (Syva, Dade-Behring, Brussels, Belgium).
Blood samples were tested for syphilis using the RPR test (Becton Dickenson, Becton-
Benelux, Erembodegem, Belgium) and for HIV-1 using ELISA Detect and Recombigent
(Cambridge-Biotech Corp., Boston, MA). The Papanicolaou smear was read at the
Department of Pathology (University of Nairobi) with quality control at the University of Ghent
(Belgium).
Data Analysis
The data were entered using Epi-info version 6.0. SPSS version 7.0 (SPSS, Inc., Chicago,
IL) for Windows was used for univariate and multivariate analyses.
PREVALENCE AND RISK FACTORS OF STI 24

Results
Demographic Data
A total of 520 women were recruited (Table 1). The mean age was 26 years (SD, 6.8; range,
14-49 years); 54% of the women were married or cohabiting, 31% were single, and 10%
were either separated, divorced, or widowed. Before the current relationship, 15% of women
had been separated or divorced and 1% had been widowed. Most women (59%) were
Protestant, 39% were Catholic, and 1% were Muslim. Approximately half of clients (48%) had
5 years to 8 years of schooling, whereas 41 % had 9 years to 13 years of schooling (i.e.
secondary education). Approximately half of women had no income.
Reproductive History and Sexual Behavior
Sixty-four percent of women reported first sexual intercourse between age 16 years and 19
years, and 11 % had sex at the age of 15 years or younger. Seventy-five percent of women
had only one sexual partner in the past 3 months and 11 % had two or more partners; 12%
had one new sexual partner in the last 3 months and 5% reported having two or more new
partners in that period. Most of the partners were circumcised (87%).
The mean number of pregnancies in this population was 2.1 and the mean number of
children born was 1.3. Twenty three percent of women had experienced a miscarriage or
abortion, whereas 9% had a history of stillbirth. At the time of the interview, 38% of women
were pregnant and 26% were using contraception, of which only 1% used condoms. Almost
half of women (47%) reported ever having used condoms.
One third of women (32%) reported a history of STDs. However, more women (44%) claimed
having sought treatment for genital infections in the past, predominantly for vaginal
discharge, followed by genital ulcers and genital warts.
Current Medical Problems
The main reasons for seeking medical care were lower abdominal pain (27%), vaginal itch
(22%), and vaginal discharge (20%) (Table 2). However, after probing for reports of
symptoms, the majority of women admitted having vaginal discharge (70%), followed by
pruritus vulvae (61%), abdominal pain (48%), and dysuria (31%).
PREVALENCE AND RISK FACTORS OF STI 25

The majority of the women were infected with C. albicans (35%), followed by T. vaginalis
(25%), and bacterial vaginosis (19%) (wet mounts were performed for 324 women only). N.
gonorrhoeae was detected in 6% of women and C. trachomatis was present in 4% of
women. Syphilis serology was positive in 6% of women and was associated with history of
stillbirth (P = 0.02). Genital warts and ulcers were observed in 6% and 12% of women,
respectively.
Table 1. Demographic characteristics, reproductive history and sexual behavior**
Characteristic
Demographic
Number or mean (±SD) Percent or range
Mean Age 24 (±6.8) 14 – 49
Currently married
Religion
280 54
Catholic 205 39
Protestant 304 59
Muslim 7 1
Other
Level of Education
4 1
None 26 5
Lower Primary 17 3
Upper Primary 249 48
Secondary 212 41
College
Occupation
16 3
Unemployed 90 17
Homemaker 144 28
Office worker 35 7
Business woman 130 25
Other
Monthly Income
121 23
None 271 52
Less than KSh* 2000 71 14
KSh* 2001-10000 175 34
More than KSh* 10000
Sexual History
3 1
Mean age at first sex 17.5 (±3.1) 8 – 27
Median no. lifetime sexual partners 3 1 – 50
Median no. new sexual partners in last year 0 0 – 9
Median no. sexual partners past 3 months 1 0 – 9
Mean no. pregnancies 2.1 (±2.0) 0 – 12
Mean no. children born
Contraceptive Use (n=323)
1.3 (±1.6) 0 – 9
None 201 62
Hormonal contraceptive 63 20
Intrauterine device 12 4
Barrier contraceptive 26 8
Surgical contraceptive 12 4
Natural Method 6 2
History of STD 165 32
Vaginal discharge 146 28
Genital ulcer disease 78 15
Abdominal pain/PID 87 17
Genital wart 3 1
** Data are from 520 female patients attending an STI clinic in Nairobi, Kenya.
* 1 US$ ≈ 60 KSh
PREVALENCE AND RISK FACTORS OF STI 26

The overall HIV prevalence was 29%. Table 2 : Reason for seeking medical care spontaneously
Among women less than 20 years,
mentioned with probing.*
Main Reason After Probing
28% were HIV positive, whereas this
n % n %
rate was 27% in the 20 years to 29 Vaginal discharge 106 20 364 70
years age group, 38% in the 30 years
Abdominal pain 141 27 249 48
to 39 years age group, and 26% in
Vaginal itch
Dysuria
114
0
22 316
159
61
31
the 40 years to 49 years age group. Fever 0 36 7
The HIV prevalence was significantly
Other 159 31 85 17
*Responses exceed 100% because of multiple problems
higher in the age group of 30 years to
39 years compared with other age groups (P = 0.04). The prevalence of STDs among
pregnant and non-pregnant women is shown in Table 3. The HIV prevalence among nonpregnant
women was 34% compared with 23% among pregnant women (P =0.01). Pregnant
women also had significantly more candidiasis (P = 0.02) and significantly less genital ulcers
(P =0.03) than pregnant women. In 22% of the patients, none of the STI pathogens could be
identified. CIN was found in 13% of women (95% CI, 9.8-15.8), of whom 8.4% had CIN I,
3.2% had CIN II, and 1.2% had CIN III carcinoma in situ.
Table 3. Prevalence of reproductive tract Infections per pregnancy status and number of sex partners***
Non-pregnant
women
Pregnancy Status No. of Sex Partners*
Pregnant women One Less than one
(n=289) (n=197) (n=463) (n=57)
n % n % n % n %
HIV 97 (34)** 23 (23)** 125 (27)** 26 (46)**
Neisseria gonorrhoeae 18 (6) 9 (5) 22 (5)** 11 (19)**
Chlamydia trachomatis 8 (3) 13 (7) 18 (4) 5 (9)
Trichomonas vaginalis 39 (24) 40 (29) 66 (23)** 16 (41)**
Candida albicans 92 (32)** 82 (42)** 166 (36) 17 (30)
Bacterial vaginosis 37 (22) 21 (15) 50 (18) 11 (28)
RPR positive 19 (7) 11 (6) 23 (5)** 7 (12)**
Genital warts 13 (5) 17 (9) 27 (6) 3 (5)
Genital ulcers 44 (15)** 18 (9)** 56 (12) 8 (14)
Cervical intraepithelial neoplasia
* past 3 months
** P < 0.05
37 (14) 23 (13) 52 (12) 9 (17)
*** not all test results were available for all women
Risk Factors for Sexually Transmitted Diseases
The relation between risk factors and the major STDs is shown in Table 4. The behavioral
risk factors significantly associated with HIV in univariate analysis were first sexual
intercourse on or before the age of 15 years, more than one partner in the last 3 months,
PREVALENCE AND RISK FACTORS OF STI 27

eing single, ever having been treated for an STD, pregnancy, an uncircumcised partner,
and ever having used a condom. After multivariate analysis including all these variables, only
the number of partners in the previous 3 months (P = 0.02) and an uncircumcised partner (P
= 0.01) remained statistically significant.
Table 4 : Odds Ratios and 95% Confidence Intervals (CI) for risk factors and sexually transmitted infections*
HIV Gonorrhea Chlamydia Syphilis CIN
Single/widow/divorced/separated 1.1 (0.7-1.8) 1.8 (0.7-4.7) 1.9 (0.7-5.4) 0.8 (0.3-2.0) 0.8 (0.4-1.5)
History of STD 1.1 (0.7-1.8) 0.9 (0.4-2.3) 0.2 (0.1-1.1) 2.1 (0.9-4.7) 2.2 (1.2-4.0)*
History of condom use 1.5 (1.0-2.3) 2.2 (0.9-5.3) 1.6 (0.6-4.1) 2.3 (1.0-5.4) 1.2 (0.7-2.2)
Younger than 20 years 0.8 (0.4-1.6) 0.9 (0.3-3.3) 2.6 (0.8-8.7) 1.6 (0.5-4.6) 1.6 (0.7-3.8)
Age at first sex < 15 years 1.5 (0.9-2.3) 0.9 (0.4-2.5) 0.2 (0.0-0.9)* 1.4 (0.6-3.5) 0.9 (0.4-1.7)
>One sex partner last 3 months 2.3 (1.1-4.8)* 5.6 (1.7-18.2)* 3.5 (0.9-13.2) 1.0 (0.3-3.2) 1.3 (0.5-3.5)
>= 1 new sex partners (3 months) 0.8 (0.4-1.7) 2.1 (0.4-3.5) 0.7 (0.2-2.9) 3.8 (1.3-10.9)* 0.9 (0.3-2.2)
Partner circumcised 0.5 (0.3-0.9)* 2.2 (0.7-6.7) 0.6 (0.2-2.0) 0.9 (0.3-2.6) 0.9 (0.4-1.9)
Pregnant 0.7 (0.4-1.1) 1.1 (0.4-2.8) 2.4 (0.9-6.7) 1.4 (0.6-3.4) 1.0 (0.6-1.9)
Bacterial
vaginosis
Trichomonas Candida
Genital
ulcer
Genital
warts
Single/widow/divorced/separated 1.9 (1.0-3.8) 1.5 (0.8-2.8) 1.1 (0.7-1.7) 1.3 (0.7-2.4) 1.4 (0.6-3.3)
History of STD 0.6 (0.3-1.2) 0.8 (0.4-1.5) 1.1 (0.7-1.8) 1.1 (0.6-2.0) 1.3 (0.6-3.2)
History of condom use 1.4 (0.7-2.5) 1.4 (0.8-2.4) 0.8 (0.5-1.1) 1.3 (0.8-2.3) 0.7 (0.3-1.6)
Younger than 20 years 0.7 (0.2-1.8) 0.8 (0.3-1.8) 0.6 (0.3-1.2) 1.2 (0.5-2.7) 3.2 (1.2-8.1)*
Age at first sex < 15 years 0.7 (0.4-1.5) 1.1 (0.6-2.1) 1.2 (0.8-1.9) 1.8 (1.0-3.3) 0.7 (0.3-1.9)
>One sex partner last 3 months 1.9 (0.7-5.1) 1.6 (0.7-3.7) 0.9 (0.4-1.9) 1.4 (0.5-3.8) 0.8 (0.2-3.2)
>= 1 new sex partners (3 months) 0.6 (0.2-1.5) 1.3 (0.6-2.9) 0.8 (0.4-1.5) 0.6 (0.2-1.5) 1.1 (0.3-3.6)
Partner circumcised 1.4 (0.6-3.3) 1.1 (0.5-2.2) 0.6 (0.4-1.0) 1.5 (0.8-3.2) 0.7 (0.3-1.8)
Pregnant 0.7 (0.4-1.4) 1.7 (1.0-3.1) 1.5 (1.0-2.3) 0.6 (0.3-1.2) 2.1 (0.9-4.9)
All Odds ratio adjusted for age, marital status, history of STD, history of condom use, age at first sex, number of sex partners
and circumcision of partner.
* P< 0.05
** Among 520 female patients, Nairobi, Kenya
The only risk factors for gonococcal infection was the number of sex partners in the last 3
months (P = 0.001). Women with chlamydial infection had significantly less history of first
sexual intercourse when younger than 16 years (P = 0.04). RPR seroreactivity was
associated in univariate analysis with the number of sex partners, having had a new sex
partner, and ever having been treated for STD. After multivariate analysis including those
variables, only having a new partner in the last 3 months remained statistically significant.
The only factor associated with CIN was history of an STD (P = 0.03). Bacterial vaginosis,
trichomonas infection, candidiasis, and genital ulcers were not associated with any of the risk
factors. Young age was strongly associated with presence of genital warts in multivariate
analysis (P = 0.007).
PREVALENCE AND RISK FACTORS OF STI 28

The association between HIV and the
STD was examined (Table 5).
Gonorrhea, bacterial vaginosis, RPR
seroreactivity, CIN, and genital warts
and ulcers were significantly associated
with HIV infection. After multivariate
analysis, all factors except syphilis
seroreactivity remained significantly
associated with HIV infection. The
presence of genital ulcers was
significantly associated with RPR
seroreactivity (P < 0.00).
Discussion
Table 5 : Association between STD and HIV Type 1 among
520 female patients, Nairobi, Kenya.
Univariate OR
(95% CI)
Multivariate OR
(95% CI)*
N. gonorrhoeae 2.2 (1.1-4.4) 3.7 (1.3-10.7)
C. trachomatis 0.4 (0.1-1.2)
RPR positive 2.6 (1.2-5.5) 2.6 (1.0-7.0)
T. vaginalis 0.9 (0.5-1.5)
C. albicans 1.1 (0.7-1.7)
Bacterial vaginosis 1.9 (1.2-3.2) 2.1 (1.1-3.9)
Genital warts 3.0 (1.5-6.2) 5.6 (2.1-15.0)
Genital ulcers 2.8 (1.7-4.8) 3.1 (1.5-6.4)
CIN 3.2 (1.9-5.6) 3.7 (1.8-7.6)
*Multivariate analysis including the sexually transmitted
disease significant in univariate analysis
Twenty-nine percent of women were infected with HIV. A cross-sectional study of the female
population attending this clinic has not been done before; hence, we cannot compare with
earlier prevalence rates. Studies in family planning clinics in Nairobi show much lower HIV
rates (Temmerman 1998). However, among women with STD-related reports in public health
facilities frequented by the same population as the one attending the STD clinic, similarly
high HIV prevalence rates have been found (Fonck 2000, Ndinya-Achola 1997). An alarming
finding in our study was that by the age of 20 years, 28% of women were HIV infected. This
is similar to findings of other studies performed in east Africa (Konde-Lule 1997). Women in
the younger age groups are infected with HIV at high rates because of a complex sociocultural
and economic background. Specific interventions for HIV prevention among the
youth are therefore urgently needed.
The prevalence of gonococcal or chlamydial infection in this study was rather low (6% and
4%, respectively) and corresponds to prevalence rates among asymptomatic persons (Colvin
1998). The prevalence of genital ulcers was 12%, but we did not study the etiologic diagnosis
of the genital ulcers. However, other studies performed in the same Nairobi clinic show that
there has been a decrease over time in the relative proportion of Haemophilus ducreyi while
the importance of herpes simplex virus as the cause of genital ulcers is increasing. In 1991,
68% of ulcers were culture positive for H. ducreyi (Fast 1982), whereas this value was
reduced to 31 % in 1997 (Malonza 1999). In the last study, 23% of cases were caused by
syphilis infection, 16% by HSV, 15% by mixed infections, and 15% were of unknown origin.
The prevalence of CIN in this population of STD patients was 13%, which is similar to the
PREVALENCE AND RISK FACTORS OF STI 29

prevalence found in a family planning clinic in Nairobi; however, family planning clients are
often perceived as a low-risk population (Temmerman 1998).
We failed to demonstrate an association between trichomoniasis and HIV infection, which
has been reported by others (Draper 1998). Although the wet mounts were done in the clinic
by the study physician, this was made impossible during several months because of power
interruptions in Nairobi. The association between chlamydia and HIV did not reach statistical
significance but showed a trend toward protection, which has been shown by others (Minkoff
1999). There is no definite explanation of this finding, but one possibility might be that we
used enzyme-linked immunosorbent assay for the detection of C. trachomatis. The more
sensitive polymerase chain reaction technique was not yet available; hence, chlamydia
prevalence is probably underreported and might have contributed to the lack of a significant
association with HIV. Also, selection bias may explain this finding. Minkoff et al. explain the
negative association by suggesting that HIV positive women have an increased condom use
and, hence, have lower STD prevalence rates. This explanation might be true in the United
States, but is certainly not the case in the setting where we worked, in which people seldom
know their HIV status.
Few women claimed to have had more than one sexual partner in the last 3 months. A 3month
period was used in this study as to minimize the recall bias. We think that most of
these women did have only one sexual partner and, hence, must have been infected by their
spouse or regular partner, as stipulated in other studies (Quigley 1997). If HIV and STD
prevention campaigns aim for a significant impact, the transmission between regular partners
will have to be addressed. This will be challenging, as women find it difficult to negotiate safe
sex with their spouse. More gender-sensitive prevention campaigns and womenempowerment
strategies will be needed. However, without strong commitment from the
government, enforcing laws to protect women against rape, sexual abuse, and violence in
their home, this aim will be difficult to achieve.
Few of the classic risk factors studied were significantly associated with genital infections.
This might be explained by the low occurrence of these risk factors in this population and,
hence, the relative lack of power to prove an association. Some trends did emerge for all
STDs, such as positive associations with factors relating to sexual behavior. This study
confirms that having multiple partners simultaneously heightens the risk for STDs more than
the number of new sex partners (Morris 1997).
PREVALENCE AND RISK FACTORS OF STI 30

Almost all women in this study reported that their partner was circumcised. Male circumcision
was associated with a reduced prevalence of HIV infection in the female partner, and this
association remained after multivariate analysis. The association between HIV infection and
circumcision in males has been demonstrated (Moses 1998), and a reduction in risk among
the female partners of circumcised men has been reported by Kapiga et al. There is a
possibility of reporting bias of circumcision status, as shown in other studies (Lilienfeld 1958).
However, in a recent study in Tanzania, misreporting was found mostly in men who reported
themselves as circumcised but who where found to be not circumcised on examination
(Urassa 1997). This may reflect a change in norms in that society that favor male
circumcision. Because the society in northwestern Tanzania is closely related to the society
in Kenya, the same might be true in this study. The women from the only tribe in Kenya that
does not traditionally practice circumcision reported their husbands to be circumcised 18%,
whereas for the other tribes the proportion was more than 90%. If there has been overreporting
of circumcision status in our study, the association between HIV and circumcision
might have been even stronger.
We found a low rate of regular condom use, which is consistent with findings in other African
settings. Some studies have been able to report a protective effect of reported condom use
for STD (Kapiga 1994). However, women who reported a history of condom use were more
likely to be infected with HIV. A possibly explanation of this finding might be that reported use
of condoms may be a surrogate marker for risky sexual behavior or extramarital sex.
Women engaging in such high-risk behavior may use condoms more often than other
women.
Women attending this clinic came from lower socio-economic groups that usually have less
access to preventive information from mass media or written materials. Few women reported
the practice of safe sex, but also reported few high-risk behaviors. Attendance to the clinic
provides a good opportunity for health education on a group and an individual basis, and this
opportunity should be seized for information, education, and counseling activities (especially
regarding safe-sex methods). Techniques to negotiate safe sex with regular partners could
be an important protective tool in this population. Program addressing the empowerment of
women in relation to their sexual behavior, especially with their regular partner, should be
strengthened to achieve more effect on prevention and control activities in the field of STD
and HIV.
PREVALENCE AND RISK FACTORS OF STI 31

3.2. Declining Syphilis Prevalence in Pregnant Women in Nairobi
since 1995: Another Success Story in the STD Field?
Summary
Published in International Journal of STD & AIDS, 1999;
10:405-8 by Temmerman M, Fonck K, Bashir F, Inion I, Ndinya-
Achola J, Bwayo J, Kirui P, Claeys P and Fransen L.
Untreated maternal syphilis during pregnancy will cause adverse pregnancy outcomes in
more than 60% of the infected women. In Nairobi, Kenya, the prevalence of syphilis in
pregnant women of 2.9% in 1989, showed a rise to 6.5% in 1993, parallel to an increase of
HIV-1 prevalence rates. Since the early 1990s, decentralized STD/HIV prevention and
control programs, including a specific syphilis control program, were developed in the public
health Facilities of Nairobi. Since 1992 the prevalence of syphilis in pregnant women has
been monitored. This paper reports the findings of 81,311 pregnant women between 1994
and 1997. A total of 4244 women (5.3%) tested positive with prevalence rates of 7.2% (95%
CI: 6.7-7.7) in 1994, 7.3% (95% Cl: 6.9-7.7) in 1995, 4.5% (95% CI: 4.3-4.8) in 1996 and
3.8% (95% CI: 3.6-4.0) in 1997. In conclusion, a marked decline in syphilis seroprevalence in
pregnant women in Nairobi was observed since 1995-96 (P < 0.0001, Chi-square test for
trend) in contrast to upward trends reported between 1990 and 1994-95 in the same
population.
Introduction
Untreated maternal syphilis infection during pregnancy will cause adverse pregnancy
outcomes in more than 60% of the cases. The risks of abortion, stillbirth, prematurity,
congenital syphilis and perinatal death are well documented in several African settings
(Watts 1984, Hira 1990, Temmerman 1992). In Nairobi, Kenya, the prevalence of syphilis in
pregnant women was 2.9% in 1989, showed a rapid rise to 5.3% in 1991, parallel to a rise in
prevalence of HIV-1 infections (Temmerman 1992). Since the early 1990s the National STD
Control Unit of the Ministry of Health has been strengthened, and decentralized STD/HIV
prevention and control programs were developed in Kenya. In Nairobi, medical staff at 10
Nairobi City Council (NCC) health centers were trained in syndrome approach and
counseling to manage sexually transmitted diseases (STDs) by joint efforts of the
government of Kenya supported by the European Commission (EC) and the Canadian
PREVALENCE AND RISK FACTORS OF STI 32

Development Agencies through the University of Nairobi. In addition, centralized and
decentralized programs aiming at reducing congenital syphilis, were set up in NCC public
health clinics with the support of the EC’s AIDS program whose strategies have always
focused on the fight against AIDS and STDs in a combined approach. The supervision,
monitoring and evaluation activities of the syphilis program are the responsibilities of a task
force consisting of representatives of the NCC, the Ministry of Health, the University of
Nairobi and the University of Ghent, Belgium (Temmerman 1993, Jenniskens 1995).
In 1992-93, an overall syphilis prevalence of 6.5% (95% CI: 6.1-6.9) was found in over
13,000 pregnant women visiting these prenatal clinics (Jenniskens 1995). In this paper we
describe the trend in syphilis in pregnant women between 1994 and 1997.
Methods
Population
Nairobi and its peri-urban area have a population of approximately 2-7 million inhabitants, of
whom the larger part live in peri-urban slums. Public health facilities fall within the jurisdiction
of the NCC with 54 health units providing maternal and child health (MCH) and family
planning (FP) services, of which 30 also render curative services. The centers are distributed
evenly in the peri-urban and urban areas and the catchment population per clinic is
estimated at 80,000.
Setting
In 1992, 10 NCC clinics were selected out of the 30 health centers with both MCH/FP and
curative components to have syphilis screening programs set up. The selection was done by
NCC in collaboration with the Task Force at the University of Nairobi and was based on
geographical distribution and catchment population.
Techniques
Syphilis serology was performed by clinic nurses, using the rapid plasma reagin (RPR) card
test (RPR, Wellcome, London, UK) because of its simplicity and low cost. Quality control
(Macrovalue, RPR card test, Becton Dickinson, Cockeysville, USA) was done by the
reference laboratory at the University of Nairobi (UoN), Department of Medical Microbiology.
RPR seroreactive women were promptly treated with a single dose of 2.4 million units of
intramuscular benzathin penicillin and advised to refer the partners for treatment. The
program was initiated in June 1992 and is still going on. Data were analyzed up to December
1997.
PREVALENCE AND RISK FACTORS OF STI 33

Training
Nurses at the antenatal clinics were trained in components of patient care, RPR testing
techniques, promoting behavior change and counseling women with syphilis and their
partners.
Equipment
Health centers’ MCH clinics were provided with laboratory equipment (centrifuge and RPR
card shaker) and supplies, drugs and counseling facilities.
Supervision
Monitoring and supervision were performed on a monthly basis using a standardized form
with checklists to evaluate the laboratory, stock management, clinic staff on duty, record
keeping, quality control and partner tracing activities.
Statistical methods
95% confidence intervals were calculated and Chi-square for trends was used to study
trends over time.
Results
A total of 81,311 pregnant women were screened for syphilis during their first visit to one of
the antenatal clinics in Nairobi, over a period of 44 months (1994-97) (Table 1). Of them,
4244 (5.2%, 95% CI: 5.1-5.4) women were found to be RPR reactive.
Table 1: Syphilis seroprevalence in pregnant women in Nairobi between 1992 and 1997.
N tests N test positive Test positive % 95% CI P for trend
1992-1993 13131 860 6.5 6.1-6.9
1994 10267 739 7.2 6.7-7.7
1995 17557 1279 7.3 6.9-7.7 0.03
1996 24480 1124 4.5 4.3-4.8
1997 29007 1102 3.8 3.6-4.0 < 0.0001
Total 81311 4244 5.2 5.1-5.4
Over an 8-month period in 1994, 10,267 were screened of whom 739 tested positive (7.2%,
95% CI: 6.7-7.7). In 1995, 1279 out of 17,557 women tested were found to be RPR positive
(7.3%, 95% CI: 6.9-7.7). Due to logistic constraints, screening was offered on alternating
days, still leaving significant proportions of the pregnant population unscreened.
PREVALENCE AND RISK FACTORS OF STI 34

From 1996, RPR tests were carried out on a daily basis covering all pregnant women
attending the primary health-care clinics.
Earlier reports from the same prenatal clinics showed a RPR seroreactivity rate of 6.5%
(95% Cl: 6.1-6.9) in 1992-93 (Jenniskens 1995). Taking these figures into account we found
a significant rise in syphilis prevalence rates between 1992 and 1996 (Chi-square for trend:
P=0.03).
In 19%, 24,480 RPR tests were
Figure 1: Syphilis prevalence (1992-1997)
performed, 1124 were found to
be seroreactive, resulting in a
9
8
syphilis prevalence of 4.5% (95%
7
Cl: 4.3-4.8). In 1997, 1162 out of
6
29,007 (3.8%) women examined,
5
4
were RPR positive (95% CI: 3.6-
3
4.0). The Chi-square test for
2
trend showed a significant
1
decrease in syphilis prevalence
0
1992-1993 1994 1995 1996 1997
between 1995 and 1997
years
(P

We believe that the observed decline in syphilis rates in pregnant women in Nairobi is a real
event. Population differences and consultation bias cannot be excluded but are not likely to
play a major role, as population characteristics (age, parity, ethnicity) have remained
unchanged over the years. The number of women screened has increased substantially
since 1995, which could be responsible for a selection bias. Most likely however, the
coverage of the program, expanding slowly, was not complete until 1995-96, and explains
the rising numbers of women in the study. Obviously, the syphilis control program by itself
might have an impact on the declining syphilis rates. The decentralized syphilis program was
initiated in Nairobi in June 1992. From that time high numbers of syphilis seroreactive women
in the reproductive age group have been adequately treated for syphilis in pregnancy. Since
the mean interval between 2 pregnancies is about 2 years for Kenyan women, some of them
may attend the clinics for a subsequent pregnancy and will not be RPR positive anymore
because of previous treatment, unless re-infection occurred.
Syphilis prevalence in pregnancy seems to show a large spread in different African countries.
In a nationwide study performed in Burkina Faso in 1994-95, a syphilis prevalence of 2.5%
was found in pregnant women (Sangare 1997). In Rwanda, in a prospective cohort study in
1992-93, 5% of the pregnant women tested positive for syphilis (Leroy 1995). Guarenti et al.
reported a syphilis prevalence of 4.8% in pregnant women in 4 different areas in Tanzania in
1995 (personal communication). A high syphilis prevalence was found in South African
publications, varying from 17% in 1990 (Bam 1994) to 9.3% in 1992 (Qolohle 1995) and
6.5% in 1994 (Wilkinson 1997). These figures could reflect a decline in syphilis prevalence in
South Africa between 1990 and 1994, but are derived from different settings, not allowing
calculating trends over time.
In conclusion, this study provides data on trends in syphilis screening results in a set
population of pregnant women in Nairobi between 1994 and 1997. Until 1995 the syphilis
prevalence was on the rise, followed by a marked decline in 1996 and 1997. Population bias
or other unrelated factors could theoretically play a role, but we believe that this decline is
real suggesting an effect of prevention and intervention programs aimed at sexual behavior
modifications, changes in health-seeking behavior and improved health services.
PREVALENCE AND RISK FACTORS OF STI 36

C HAPTER 4
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL
BEHAVIOR IN NAIROBI
4.1. Healthcare-Seeking Behavior and Sexual Behavior of Patients
with Sexually Transmitted Diseases in Nairobi, Kenya
Summary
Published in Sexually Transmitted Diseases, 2001; 367-371 by
Fonck K, Mwai C, Rakwar J, Kirui P, Ndinya-Achola J, and
Temmerman M.
Sexual and health-seeking behaviors are important components of sexually transmitted
disease (STD) control. We undertook this study to generate data for improved STD
prevention and care, and to assess sexual behavior and relevant health-seeking behavior.
A questionnaire to elicit social, demographic, healthcare-seeking, and sexual behavior
information was administered to 471 patients attending the referral clinic for STDs in Nairobi,
Kenya. A large proportion of the patients had sought treatment in public and private sectors
before attending the clinic for STDs. Women waited longer than men to seek medical care.
In addition, women more than men engaged in sex while symptomatic, mostly with their
regular partner. Condoms were used rarely during illness. In their self-reports, 68% of the
men admitted to having extramarital affairs, and 30% to paying for sex, yet they blamed their
wives for their STDs. Health education messages in Kenya need adaptation to improve
health-seeking behavior and safe sex practices.
Introduction
In Kenya, sexually transmitted diseases have long been recognized as a major public health
problem because of their high prevalence and significant contribution to morbidity and
mortality (Piot 1990, Hunter 1994, Temmerman 1990 & 1998). Furthermore, it is known that
STDs, both ulcerative and non-ulcerative, facilitate HIV-1 transmission (Plummer 1991,
Chapter 4 37

Wasserheit 1992, Laga 1993). In Nairobi, STDs represent 5% to 10% of the caseloads at
many outpatient clinics (Reach 1988). Although the overall prevalence of STDs in the
general population is not known, studies among specific population groups have illustrated
high STD prevalence in both low- and high-risk groups.
Work performed in 1989 among pregnant women in Nairobi showed prevalence rates of 6%
for HIV, 3% for Neisseria gonorrhoeae, and 2% for syphilis (Temmerman 1992). Among
women attending a family planning clinic in Nairobi, the prevalence rate was 10% for HIV-1,
4% for Chlamydia trachomatis, 2% for N. gonorrhoeae, 2% for syphilis, and 1% for genital
ulcer disease (GUD) (Temmerman 1998 and 1995). A 1991 study of men attending an STD
clinic in Nairobi showed an HIV seroprevalence of 24% (Tyndall 1994) among men with nonspecific
symptoms as well as prevalence rates of 37% for N. gonorrhoeae and 33% for C.
trachomatis (Tyndall 1999).
The effectiveness of STD prevention and treatment programs depends, among other factors,
on adequate knowledge about the health behavior and health-seeking behavior of the
population, and on the sexual behavior characteristics of the target group. It is important to
understand the dynamics and interactions that make a person decide why, when and where
to seek care, and which health care facility to attend. To provide adequate health care, it is
imperative to know whether the choice of the health facility for an STD differs from the choice
for other common health problems. Issues such as accessibility and acceptability of services
are key to the provision of successful STD services.
This study examined the reasons for and patterns of health-seeking and related sexual
behaviors of patients with STD-related signs and symptoms who present at STD referral
clinics. The objective was to identify health-seeking behavioral factors and health servicerelated
issues that can be used to improve STD services, and to develop education
campaigns for the general public about seeking timely and correct STD treatment.
Methods
This study was conducted at the main public health STD referral clinic in Nairobi, also known
as the Special Treatment Center. The clinic is located downtown and draws its patients from
public and private institutions both within and outside Nairobi. The clinic, operated by the
Nairobi City Council, serves as a referral clinic for approximately 50 primary health care
clinics in Nairobi. Most of the peripheral public health clinics in Nairobi do not provide STD
care, but reroute the patients with STD-related symptoms to the STD clinic.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 38

Most of the patients are low-income earners. The clinic has a well-equipped laboratory with a
staff highly trained in STD management.
Between February and May 1998, a convenience sample of new patients was interviewed by
trained research assistants who approached the patients leaving the clinic. Whoever
happened to be seen leaving the clinicians' room or pharmacy was asked to participate in the
study. After giving verbal informed consent to participate, patients were interviewed using a
detailed pretested questionnaire covering social and demographic information, patterns of
health-seeking behavior, and sexual behavior. The questionnaires were linked later to the
medical records. The questionnaires were available in both English and Kiswahili, so the
interviews were conducted in the language most comfortable to the patient.
All the survey responses were entered into a computerized database using SPSS for
Windows, version 7.0 (SPSS, Chicago, IL). Cross-tabulation and descriptive statistics were
calculated using χ 2 tests for categorical variables.
Results
Most of the patients recruited for the study accepted to participate, and 471 people were
interviewed. Only 21 patients refused Participation, mainly because of time constraints.
Table 1 summarizes the demographic characteristics of the study group. Half of the
participants were women, who were younger and more often married than the men. There
was no difference in education level between the genders, but almost half of the women
reported not having any income, as compared with only 11% of the men.
The participants were asked what they would do if they had a general health problem.
According to their self-reports, 57% of the women would prefer public health services, as
compared with 57% of the men, who more often would opt for private facilities. Mission
hospitals and chemists were mentioned less often as a first choice. Traditional healers would
be visited by only two of the men, and by none of the women.
The choice of a particular health facility for general health care was determined mainly by
convenience of the location (men 56%, women 55%), privacy (men 18%, women 17%), and
affordability (men 12%, women 18%) of the services.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 39

Table 1: Demographic data of 471 patients attending an STD clinic in Nairobi, Kenya.
Male (n=234)
n or mean (% or range)
Female (n=237)
n or mean (% or range)
P-value
Age 28.4 (17-53) 24.8 (16-55) < 0.000
Marital status 0.02
Married 119 (51) 144 (61)
Single 107 (46) 82 (35)
Separated/divorced/ widowed 8 (3) 11 (5)
Educational level NS
None 4 (2) 4 (2)
Primary 90 (39) 113 (48)
Secondary 115 (49) 106 (45)
Post secondary 25 (11) 13 (6)
Income < 0.000
None 25 (11) 108 (46)
< 1000 KSh* 10 (4) 15 (6)
1000-5000 KSh 126 (54) 89 (38)
> 5000 KSh
* 1 US$ ~= 60 KSh
NS = not significant
73 (31) 25 (11)
Patients were further interviewed on their health-seeking behavior related to the actual STD
episode. The public health sector was less attended for this STD episode than it would be for
general health problems, and this was more obvious for the men. Instead, the participants
went more often to chemists, traditional healers, private hospitals, and mission hospitals.
The women did not significantly change their preference in relation to the type of health
problem. More of the men came to the STD clinic directly, whereas the women visited other
health facilities before the STD clinic (67% versus 55%, P = 0.01).
Furthermore, women more than men had been referred from other health facilities to the
STD clinic (44% versus 26%, P < 0.000) (Table 2). Most of the referred patients, both men
and women, had attended another Nairobi City Council clinic before coming to the STD
clinic. Few patients were referred from private clinics.
Most of the participants had spent considerable amounts of money on treatment for their
illness before attending the STD clinic, averaging 1196 shillings (KSh) (~20 US$). The men
had spent twice as much as women on treatment before attending the STD clinic (mean,
1582 KSh versus 880 KSh; median 600 KSh versus 400 KSh). Mission hospitals were the
most expensive.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 40

Table 2: Treatment sought elsewhere and referral status among women and men attending the STD clinic.
Men Women
Had Sought Treatment Had Been Referred Had Sought Treatment Had been Referred
Before (n=150)
(n=61)
Before (n=187)
(n=104)
Type of facility
n (%)
n (%)
n (%)
n (%)
Nairobi City Council 36 (24) 45 (74) 79 (42) 80 (77)
Government 14 (9) 2 (3) 18 (10) 12 (12)
Private 71 (47) 7 (1) 65 (35) 8 (8)
Chemist 13 (9) 0 (0) 5 (3) 0 (0)
Mission 11 (7) 3 (5) 19 (10) 3 (3)
Traditional healer 5 (3) 0 (0) 1 (1) 0 (0)
Other 0 (0) 4 (7) 0 (0) 1 (1)
Of the men, 181 (77%) were found to have urethritis, 12 (8%) genital ulcer disease, 6 (3%)
other infections, and 19 (8%) no STD. Of the women, 83 (35%) were found to have vaginitis,
95 (40%) cervicitis, 30 (13%) pelvic inflammatory disease, 12 (8%) genital ulcer disease, 5
(2%) other infections, and 15 (6%) no STD.
Table 3 shows the delay between the appearance of symptoms and the seeking of medical
care by gender and by medical problem. The participants were symptomatic for an average
of 1 week before seeking treatment, but the range of delay was wider for the women.
Moreover, the participants, particularly the women, postponed their visit to the STD clinic for
approximately 1 month, except in case of genital ulcer disease. Referred patients had a
significant shorter delay before attending a health facility than the non-referred patients (12.7
days versus 40.6 days, P = 0.01).
Table 3: Delay in days (median and percentiles [25-75]) between appearance of symptoms and medical contact, by gender.
Days Before Seeking Treatment
Days Between First Treatment and Visit to
STD clinic
Syndromes Man Women Man Women
Any complaint 7 (3-14) 7 (3-21) 23 (5-60) 29 (7-90)
Urethritis/PID/cervicitis 7 (3-14) 7 (3-30) 28 (5-60) 27 (7-90)
GUD 7 (3-14) 7 (3-14) 14 (7-30) 16 (7-30)
Vaginitis 7 (3-21) 46 (7-90)
Non-STD 7 (2-21) 14 (3-30) 16 (7-60) 29 (3-150)
Table 4 shows the participants' sexual behavior characteristics. Approximately half of the
patients, significantly more women than men engaged in sex while symptomatic. Most of the
women had sex with only one partner, who in most cases was their spouse or a regular
partner. About one third of the men who engaged in sex while symptomatic had sex with two
or more partners, and 22% had casual sex partners or sex in exchange for money. Only 19%
of those who engaged in sex during illness used condoms, the men more often than the
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 41

women. Overall, significantly fewer women reported ever having used condoms (48%
versus 63%, P = 0.001).
Significantly more men (45%) than women (30%) had a history of STD (P < 0.000).
Participants with a history of STD reported significantly more extramarital sex (54% versus
22%, P < 0.000) and had been referred less often (29% versus 39%, P = 0.05). More men
than women had attended the STD clinic before (34% versus 24%, P = 0.01). Patients who
had attended the STD clinic earlier or more often reported multiple partners (24% versus
12%, P = 0.04), previous STDs (65% versus 26%, P < 0.000), and sex in exchange for
money (8% versus 2%, P = 0.04).
Table 4: Sexual behavior during illness of 471 patients attending the STD clinic in Nairobi, Kenya.
Men
n %
Women
n %
Sex while symptomatic 79 (34) 146 (62)
P
< 0.001
Number sex partners while ill:
1 54 (68) 137 (94) < 0.001
2 or more
Sex partner
25 (32) 9 (6) 0.006
Spouse 42 (53) 97 (66) < 0.001
Girlfriend/boyfriend 37 (47) 22 (15) 0.04
Casual partner 7 (9) 4 (3) NS
Person who received/gave gifts or money for sex 10 (13) 2 (1) 0.04
Condom use while symptomatic
Condoms used with partner
17 (22) 26 (18) NS
Spouse 5/17 (29) 15/26 (58) 0.04
Girlfriend/boyfriend 8/17 (47) 8/26 (30) NS
Casual partner 2/17 (12) 2/26 (7) NS
Person who received/gave gifts or money for sex
NS= not significant.
4/17 (24) 0 NS
As to the source of the current STD, only 57 (24%) of the men and 8 (3%) of the women
responded. Men blamed mainly their regular partner for having contracted an STD, despite
more self-reported risky behavior (Table 5).
Most of the married men did not suspect their spouse of extramarital affairs, whereas almost
half of the women thought their husbands engaged in extramarital sex. This was supported
by 68% of the men self-reporting extramarital affairs, as compared with only 6% of the
women. The unmarried men with a regular partner more often had a devious suspicion that
their girlfriend had other partners. One third of all the men in the study also reported that they
had paid for sex, as compared with 7% of women who admittedly had received money in
exchange for sex.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 42

Table 5: General sexual behavior among individuals with a regular partner attending the STD clinic.
Married Partners
Partners With Regular
Boyfriend or Girlfriend
Men Women
Men Women
(n=120) (n=144) P (n=76) (n=57) P
n (%) n (%)
n (%) n (%)
Do you think your spouse has extra-marital affairs:
Yes 6 (5) 69 (48) =3
NS= not significant
9 (13) 1 (1) 0.006 4 (5) 3 (4) NS
Discussion
One parameter determining the transmission dynamics of STDs in a given population is the
duration of time an individual with an STD is sexually active (Anderson 1987). The lack of
successful treatment increases the duration of infectiousness, thus contributing to STD
transmission and complications. Hence, the impact of an STD program depends, besides the
effectiveness of the treatment regimens, on the time that infected individuals wait before
seeking treatment and the extent of their unprotected sexual activity during this period. This
study examined a cross-section of patients attending the STD referral clinic in Nairobi.
These individuals constitute a high-risk group for transmitting STDs, including HIV/AIDS, and
may be a priority target for interventions.
The patients in the current study had waited one week, on the average, before seeking
medical care. These findings are consistent with an earlier study conducted in Nairobi among
STD patients in primary health care clinics (Moses 1994). Various reasons have been
advanced for this long waiting time: (1) Lack of knowledge about the importance of seeking
prompt medical care may play a role, which may be reflected by the extremely long delay
before seeking care by women with pelvic inflammatory disease or cervicitis. (2) Lack of
available services in the Nairobi area could be another factor. More than half of the patients
in this study reported convenience of the location as the most important factor explaining why
they attended a specific facility. (3) Finally, lack of financial resources, especially for women,
also might play an important role in causing delayed treatment. Most of the women in this
study reported not having an income. They, therefore, are dependent on their partners for
money. In addition, findings show that women in this society are apprehensive about their
partners' reaction on discovering that they have acquired an STD (Temmerman 1995).
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 43

Stigmatization by health staff also may play a role in preventing some patients from seeking
medical care.
In addition, the delay between attending a primary health facility and the STD referral clinic
was considerable, which is quite worrisome. More than half of the patients had sought
treatment elsewhere before attending the STD clinic. Treatment for STDs is not available at
all primary health care clinics in Nairobi. Therefore, patients with STDs often must be
rerouted immediately to the nearest clinic with STD treatment or to the STD referral clinic.
Consequently, the clinics with STD treatment should refer only the patients with STD who do
not respond to the first-line syndromic treatment approach provided.
This system, although still causing some delay in appropriate treatment, seems to function
well. However, more operational research on the efficiency of referral is needed. The delay in
presenting to the STD clinic is considerably shorter for referred patients, indicating that
referral status decreases the stigma attached to attending an STD clinic. Patients who have
been to the STD clinic before find it much easier to come straight to the STD clinic for the
next episode.
With regard to health-seeking behavior, men more often than women report a different
healthcare-seeking attitude when confronted with an STD. The men in this study reported
attending the public health sector significantly less often for an STD than for other health
problems. This implies that adequate STD case management must be offered as widely as
possible to have a significant impact on the STD epidemic.
The women were more likely than the men to have unprotected sex while symptomatic, the
married women more so than the unmarried women. This observation further reinforces the
belief about the imbalance in African families, with the woman possibly unable to refuse sex
or negotiate safe sex with her husband or regular partner. Thus, the women, expected to be
sexually and socially subordinate to men, continue to be infected by their partners (Ickovics
1998). This observation is supported by the fact that two thirds of the men in the current
study admitted their extramarital sex activity, but still blamed their wives for contracting the
STD (Rakwar 1999).
A similar observation was made by Moses et al. in 1994, who found that married women
were more likely than unmarried women to engage in sex while symptomatic, and that a
large proportion of them had sex several times before seeking treatment. Men also reported
more multiple partners than women. Therefore, men are more likely than women to transmit
STDs to their partners. This is consistent with the belief that African women are at increased
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 44

isk for STD/HIV infection through their spouses (Hunter 1994, Ulin 1992, Quigley 1997). It is
therefore very important to educate men on taking their responsibilities, by either changing
their behavior or using safe sex methods. Hence, health education at various levels must be
improved, and messages should be directed more specifically to men.
One third of the men in this study admitted to having ever paid for sex with money or gifts.
Because of the economic situation in Kenya, many women need extra income, be it for rent
or school-fees, which they obtain through one or more regular boyfriends in exchange for
sex.
The current study shows that some men and women are at high risk for STDs, including HIV
infection, through large numbers of sexual partners. Intervention strategies therefore should
aim at reducing partner change and promoting condom use. However, some individuals with
a low-risk profile, especially women, may be at risk through their spouses or regular partners.
This suggests that interventions should extend beyond the high-risk groups, and men
especially should be targeted. Simultaneously, however, the quality of existing primary health
care services should be strengthened, with STD treatment extended to more, if not all,
primary health care clinics in Nairobi. These services should include health education that
especially targets prevention strategies.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 45

4.2. Health-Seeking Behavior and Sexual Behavior among Primary
Health Care Patients in Nairobi, Kenya.
Summary:
Published in Sexually Transmitted Diseases 2002; 29:106-111,
by Fonck K, Mwai C, Ndinya-Achola J, Bwayo J and
Temmerman M.
Health-seeking and sexual behaviors are important elements in the control of sexually
transmitted infections (STIs). We examined patterns of health-seeking behavior and related
sexual behavior relevant to improved prevention and care among patients attending primary
health care (PHC) clinics. A Questionnaire covering social, demographic, and healthcareseeking
and sexual behavior information was administered to 555 patients attending three
primary health care clinics in low socio-economic areas or Nairobi, Kenya. Women's
knowledge about health in general and STIs in particular was poor. A major gender
difference in delay of health-seeking for STIs was observed (5 days for men versus 14 days
for women). Significantly more men than women reported a history of STIs (68% versus
47%; P = 0.04). Men reported more extramarital affairs (17% versus 8%; P < 0.001). A high
prevalence of gonorrhea (3%) and chlamydia (6%) was found in this population, with no
difference between the genders. The urine dipstick was ineffective for the detection of these
STls. There is a need for better understanding of behavioral factors, as well as gender and
social aspects of health care. Health education and health promotion in these areas should
be strengthened. Improved screening tests are needed for the detection of STIs.
Introduction
Sexually Transmitted Diseases (STDs) have long been recognized as a major public health
problem in Kenya because STIs and their sequelae contribute significantly to morbidity and
mortality in the population (Piot 1990, Ryder 1991, Berkowitz 1993, Temmerman 1992). In
addition it is known that most STIs facilitate HIV-1 transmission (Wasserheit 1992,
Sewankambo 1997, Schacker 1998, Kreiss 1994, Cameron 1989, Laga 1993). In Nairobi,
STIs represent between 5 and 10% of caseloads at many outpatient clinics (Reach 1988).
The overall prevalence of STIs in the general population, however, is unknown, but studies
among specific population groups have illustrated high STIs prevalence in both high- and
low-risk groups.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 46

Work performed in 1989 among pregnant women in Nairobi showed a prevalence of 6% for
HIV, 3% for N. gonorrhoeae, and 2% for a positive syphilis serology (Temmerman 1992).
Among women attending a family planning clinic in Nairobi, the prevalence was 10% for HIV-
1, 4% for Chlamydia trachomatis, 2% for N. gonorrhoeae, 2% for syphilis, and 1% for genital
ulcer disease (Tyndall 1998).
An important determinant of the transmission dynamics of STIs is the duration of the infection
in a sexually active individual (Brunham 1990). A survey in Nairobi showed that patients were
symptomatic for an average of one week before seeking treatment, but the range of delay
was wider for women (Fonck 2001). In addition, 62% of women versus 34% of men had sex
while symptomatic. These findings were similar to those from an earlier study in Kenya
showing that 42% of symptomatic patients had waited one week before seeking treatment,
and 23% delayed two weeks (Moses 1994).
The development of effective STI treatment and prevention programs depends on adequate
knowledge about health-seeking behavior and related sexual behavior in the population. It is
important to understand the dynamics and interactions that make an individual with a health
problem choose a certain health care facility. To provide adequate health care, it is
imperative to know whether the choice of health facility in the case of an STI is different from
that for other common health problems. Issues such as accessibility and acceptability of
services are key to providing successful STI services.
In this study, we examined the reasons for and patterns of health-seeking and related sexual
behavior among patients attending primary health care clinics in Nairobi. The objective was
to identify health-seeking behavioral factors and health service-related issues that can be
used to improve STI care, and to develop education campaigns for the general public about
seeking timely and correct STI treatment.
Methods
The study was carried out at three primary health care clinics (PHC) in different areas of
greater Nairobi. All three operated by the Nairobi City Council (NCC). Two of the three clinics
provide STI services and are located near large slum areas. The third clinic was chosen
because it is one of the few NCC clinics open on a 24-hour basis. The majority of the
patients attending these clinics are low-income earners. Treatment in public health facilities
is not free of charge, but costs considerably less than in the private sector.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 47

Between September and November 1998, a convenience sample of 244 men and 311
women was interviewed by trained research assistants, who approached the patients while
they were leaving the clinic. Whoever happened to be seen leaving the clinician’s room or
the pharmacy was asked to participate in the study.
After the individuals gave verbal informed consent, exit interviews were conducted using a
detailed pretested questionnaire covering social and demographic information, patterns of
health-seeking behavior, and sexual behavior. The questionnaires were available in both
English and Kiswahili, with interviews conducted in the language most comfortable to the
patient.
Urine samples collected from the patients were transported to the Department of Medical
Microbiology, University of Nairobi, where a polymerase chain reaction (PCR) test for
detection of Neisseria gonorrhoeae and Chlamydia trachomatis was performed using the
Amplicor PCR Diagnostics (Roche Diagnostic System, Ontario, Canada). Clinical patient
card data were correlated with study variables.
Data Analysis
All survey responses were entered into a computerized database using Statistical Package
for Social Sciences (SPSS) for Windows, version 7.0 (SPSS, Chicago, IL). Cross-tabulation
and descriptive statistics were calculated using Chi-square tests for categorical variables.
Results
On the 555 patients interviewed, 311 (56%) were women. Table 1 shows the demographic
characteristics of the patients. The mean age of women (27 years) was lower than that of
men (30 years), (P

The reason for this choice was convenience of clinic location (32%), followed by quality of
services (26%) and cost of services (12%). Confidentiality or friendliness of the staff was
rarely mentioned as a reason for the choice.
Significantly more men than women had ever been treated for an STI. For both genders, the
healthcare-seeking behavior related to STI was different from that related to general health
problems. Public health facilities and mission hospitals were rarely visited for an STI episode.
Men attended private clinics significantly more often with an STI than with general health
problems.
Table 1: Demographic characteristics of 555 patients attending primary healthcare clinics in Nairobi, Kenya.
Men Women
(n=244)
n
%
or range
(n=311)
n
%
or range
P-value
Age
Marital status
30 16-70 27 14-75

Table 2: Healthcare-seeking behavior related to general health problems and to Sexually Transmitted Diseases (STDs) among
Nairobi City Council clinic attendees.
Men
Women
(n=244) (n=311)
n % n % P- value
What is the most common health problem where you live?
Diarrhea, cholera, amoebiasis
24 10 23 8
Cold, cough, pneumonia
48 20 64 21
Malaria
99 41 91 30
STD/HIV/AIDS
11 5 11 4
Do not know
Where do you go for most of your health problems?
32 13 81 27

The participants were further
interviewed about their actual health
problem (Table 3). Significantly
more women were referred to the
clinic. About one third of the patients
were attending the clinic for the first
time, and significantly more women
than men had visited the actual
clinic in the past. The majority had
walked to the clinic, and this in most
cases had taken less than one hour.
Men were served significantly faster
in the clinic than women, with 70%
spending less than one hour in the
clinic. It must be noted that the clinic
cards did not indicate any health
problem or any diagnosis for 26% of
the men and 52% of the women.
Table 4: Delay in days between appearance of
actual symptoms and medical care-seeking, by
gender.
Complaints Male Female P
Headache
Fever
Cough
Diarrhea
Injuries
Stomach pain
Throat pain
STI
Other
Not indicated
8.7
6.0
6.0
3.0
6.0
16.8
7.5
4.9
9.5
10.4
8.6
5.7
9.5
8.3
3.4
28.3
3.0
13.7
11.8
13.7
0.04
0.05
Table 3: Health-seeking behavior in relation to actual episode of illness.
Men Women
n % n % P
Referred
Times been to this clinic, including today
54 22 90 29 0.04
1
74 31 90 29
2
55 23 49 16 0.05
3-10
79 33 108 35
>10
Last time been to this clinic:
30 13 73 24 0.001
< 1 week
19 18 22 13
1-4 weeks
10 10 27 16
> 4 weeks
76 72 119 71
Walked to clinic
Time to reach clinic:
198 83 244 79
3 hours
Time spent in clinic:
4 2 5 2
< 1 hour
170 70 189 61
1-2 hours
62 25 103 33
> 2 hours
12 5 19 6
The delay between appearance of symptoms and
seeking medical care varied widely, from 3 days for
an episode of diarrhea among men to 28 days for
stomach pain among women (Table 4). Although
women in general waited longer before seeking
care, there was no significant difference in delay for
most disease episodes, except for diarrhea and
reports of STI. Of the 29 patients with recorded
reports of STI, men had waited an average of 5 days
before coming to the clinic, as compared with 14
days for women (P=0.05).
The sexual behavior of the clients is shown in Table 5. The majority of the clients reported
having a regular partner, but significantly more women also lived together with their partner.
Significantly more women said their partner also had other sex partners, whereas
significantly more men said their partner was faithful. Only 17% of men admitted extramarital
affairs, which was significantly higher than among the women. Of the men who had affairs,
75% had engaged in sex with two or more partners in the last 6 months.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 51
0.04

Thirteen women had engaged in sex with more than one new partner in the last 6 months.
Condom use with irregular partners was rather low, with no difference between men and
women (38% and 29% respectively; P=0.7).
Table 5: Sexual beliefs, behavior, and history associated with treatment of STDs in 555 patients attending primary health care
clinics.
Men Women
n % n %
mean or range mean or range P
Boys become sexually active 14 7-28 14 7-25
Girls become sexually active 12 6-25 13 7-20
Ever had sex 231 95 277 89 0.02
Age at first sex 17 10-30 18 10-28
Have regular partner 198 86 230 83
Live with regular partner 85 43 171 74

Of the patients interviewed, 68 (12%) refused to give a urine sample, with no statistical
difference between the men and women (10% and 14% respectively, P=0.08). The dipstick
results were positive in 109 (29%) of the patients, and no association with STI-related signs
and symptoms was observed. Overall, 12 (3%) samples yielded positive PCR results for
gonorrhea, 28 (6%) for chlamydia, and 2 for both. No difference in prevalence between the
genders was observed. A positive PCR test result for gonorrhea, chlamydia or both was not
associated with current reports of STI or with a history of STI or any other risk factor. Of the
persons with positive PCR results for chlamydia, 9 (32%) had no reports of health problems
on their medical cards. This was the case for 7 (58%) patients with positive PCR results for
gonorrhea. None of the clinic cards of patients with positive PCR results mentioned any STI
report, symptom, or diagnosis.
The sensitivity, specificity, and positive predictive value of the dipstick for the detection of
gonorrhea were 50%, 78%, and 6% respectively. For chlamydia those values were 26%,
88%, and 14% respectively.
Discussion
We found important gender differences in knowledge about health in general, and in healthseeking
behavior among primary health care patients in a lower socio-economic population
in Nairobi. This finding is similar to that described in other studies from both developing and
developed countries, indicating that various socio-cultural and gender factors contribute to
the delay in health-seeking (Johansson 2000, Oberlander 2000, Meyer-Weitz 2000). A study
from Bangladesh reported that women with illness seek care significantly less often than men
(Ahmed 2000). Patients may also neglect symptoms until the disease reaches a serious
stage before seeking medical care. Furthermore, the kind of symptoms may influence the
likelihood that patients will seek help (Goldman 2000).
In western Kenya important differences in self-treatment practices and choice of medicines
between boys and girls were found (Geissler 2000), which may reflect the higher income
potential of boys. In Nairobi, the lack of economic means was found to be the most important
factor influencing women with STI symptoms not to seek care (Fonck 2001). This may have
serious consequences because it has been shown that the duration of sexually activity is one
of the parameters determining the transmission dynamics of sexually transmitted infections in
a given population (Anderson 1987).
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 53

Men reported more extramarital affairs and more histories of STI than women. Both genders
reported low condom use. Rather high prevalences of STIs were found in this population, but
with no difference between the genders. This may indicate that men are more likely to
transmit STIs to their partners than women. This is consistent with the belief that African
women are at increased risk of STI and HIV infections through their spouses (Hunter 1994,
Quigley 1997, Ulin 1992). It is therefore urgent to educate men about taking their
responsibilities and to either change their behavior or use safe sex methods. Hence, health
education at various levels in Kenya must be improved, and messages should be directed
more specifically to men.
Both men and women were found to use the public health services less often than private
sector health care when they have an STI. Another study in eastern Africa found that
dissatisfaction with state medical provision is not manifested as rejection of the allopathic
medicine with which it is associated, but as increased reliance on an emerging informal
sector of private medical provision (Green 2000). In Kenya there has been a recent marked
increase in the number of private clinics, with little control as to the quality of services they
offer. There are no data on numbers of patients treated or the proportion of patients
successfully treated. The consequences of patients receiving either inappropriate or
insufficient drugs are dangerous (Moses 1992). The public health authorities in Kenya need
to investigate the quality of treatment offered in these facilities, and if necessary, to provide
appropriate training in the management of STIs for this cadre of health workers. Although STI
management is provided in the public health sector, there exists a need to reexamine why
there is a high proportion of treatment failures in these institutions.
An important finding of our study was that the clinicians rarely recorded reports of health
problems, signs, or symptoms of STI on clinic cards, although two of the three clinics offer
integrated STI services. A possible explanation is that few patients, even if symptomatic,
reported the STI directly. Even if they mentioned it, the clinicians were reluctant to indicate it
on the cards. The training of health staff should put more emphasis on the importance of a
good anamnesis and the correct recording of findings, treating an STI as any other health
problem.
We cannot conclude from this study whether individuals with positive laboratory testing
showed signs or symptoms of STI because the cards did not indicate STI symptoms. Hence
we can assume that a mixture of both symptomatic and non-symptomatic patients presented
at the clinic. The performance of the dipstick procedure as a screening test for the detection
of gonorrhea or chlamydia was unsatisfactory in this population.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 54

In conclusion, there is a need for better understanding of behavioral factors, and for the
development of strategies that take these into account. Health workers need a better
understanding of the gender and social aspects of health care in general and STI control in
particular, especially the aspects that influence the likelihood for achieving equity in
diagnosis and cure. Furthermore, there is a clear need for health education and health
promotion on the urgency for seeking timely and correct medical care, taking into account the
different attitudes and behaviors of men, women, and health care workers.
HEALTHCARE-SEEKING BEHAVIOR AND SEXUAL BEHAVIOR IN NAIROBI 55

C HAPTER 5
DIAGNOSIS AND MANAGEMENT OF SEXUALLY
TRANSMITTED INFECTIONS
Validity of the Vaginal Discharge Algorithm among Pregnant and
Non-pregnant Women in Nairobi, Kenya.
Summary
Published in sexually Transmitted Infections 2000; 76:33-8 by
Fonck K, Kidula N, Jaoko W, Estambale B, Claeys P, Ndinya-
Achola J, Kirui P, Bwayo J and Temmerman M.
The aim of the study was to evaluate the validity of different algorithms for the diagnosis of
gonococcal and chlamydial infections among pregnant and non-pregnant women consulting
health services for vaginal discharge in Nairobi, Kenya. Cross sectional study among 621
women with complaints of vaginal discharge in three city council clinics between April and
August 1997. Women were interviewed and examined for symptoms and signs of sexually
transmitted infections (STIs). Specimens were obtained for laboratory diagnosis of genital
infections, HIV, and syphilis. The data were used to evaluate the Kenyan flow chart as well
as several other generated algorithms. The mean age was 24 years and 334 (54%) were
pregnant. The overall prevalence rates were: 50% candidiasis, 23% trichomoniasis, 9%
bacterial vaginosis, 7% gonorrhea, 9% chlamydia, 7% syphilis, and 22% HIV. In nonpregnant
women, gonococcal and chlamydial infection was significantly associated with (1)
demographic and behavioral risk markers such as being single, younger than 20 years,
multiple sex partners in the previous 3 months; (2) symptom fever; and (3) signs including
presence of yellow or bloody vaginal discharge, cervical mucopus, cervical erythema, and
friability. Among pregnant women only young age, dysuria, and fever were significantly
associated with cervical infection. However, none of these variables was either sensitive or
specific enough for the diagnosis of cervical infection. Several algorithms were generated
and applied to the study data. The algorithm including risk markers performed slightly better
than the current Kenyan algorithm. STIs form a major problem in the Nairobi area and should
be addressed accordingly. None of the tested algorithms for the treatment of vaginal
Chapter 5 56

discharge would constitute a marked improvement of the existing flow chart. Hence, better
detection tools for the specific etiology of vaginal discharge are urgently needed.
Introduction
Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) are two common causes of
genital tract infections that have a major impact on health, particularly of women and
neonates in developing countries. These infections are also known to facilitate the sexual
transmission of human immunodeficiency virus (HIV) (Kreiss 1994, Laga 1991, Plummer
1991). Therefore, sexually transmitted disease (STD) control activities not only prevent
complications but also offer an additional strategy for the prevention of HIV (Laga 1994,
Grosskurth 1995, Mayaud 1997).
As laboratory detection methods for genital infections are expensive and not widely available
in developing countries, diagnostic algorithms based on clinical signs and symptoms have
been proposed by the World Health Organization (WHO) as a tool for better management of
patients presenting with genital tract problems at the primary health care level. The
algorithms for genital ulcer disease as well as for urethral discharge have been successfully
adapted for use in different countries. The flow chart for vaginal discharge, however, poses
problems owing to the number and diversity of the pathogens. To address this problem, a
simple risk score for the identification of NG and/or CT infections in women with complaints
of vaginal discharge has been developed (WHO 1993): hence, this would result in a
reduction of overtreatment costs as well as the occurrence of side effects.
In Kenya, the algorithms for STD treatment have been derived from the WHO guidelines and
locally adapted by an expert committee. The current vaginal discharge algorithm has been in
use for several years. At the initial visit, a woman with vaginal discharge without abdominal
pain receives treatment for vaginal pathogens while treatment for pelvic inflammatory
disease (PID) is reserved for accompanying abdominal pain and tenderness (Fig 1). Upon
follow up after 7 days, women previously treated for vaginal conditions are treated for
cervical infections if there is no improvement. No risk scores or genital examination are
included in the flow chart.
Following this flow chart, women with cervical infections are systematically missed at the first
visit unless abdominal pain is present. In addition, many of the women with NG/CT infection
might be lost to follow up especially if their symptoms have improved as a result of the
vaginal infection treatment received. On the other hand, many pregnant women with
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 57

discharge and abdominal pain-two conditions often occurring in pregnancy-might
unnecessarily be treated for cervical infections at the first visit. From personal
communications with health care workers it became obvious that they often rely on their
personal (clinical) judgment to treat women for cervical infection at the first visit. To this effect
they seem to take into consideration risks, signs, and symptoms to make a diagnosis.
Hence, very often this algorithm defeats its public health goal, the control of cervical
infections, and risks of undermining the credibility of the services.
We thus undertook this study to validate the use of the current Kenyan clinical algorithm for
vaginal discharge, which is one of the most frequent reasons for consulting the health
services. We validated addition of risk scores as well as inclusion of signs and symptoms
among pregnant and non-pregnant women.
Patients and methods
Data Collection
The study was conducted between April and August 1997 at two peripheral health centers
(PHC) and at the major STD referral clinic (STC) run by the Nairobi City Council (NCC). The
contact with the patient took place in Kiswahili. All women with spontaneous or prompted
complaints of vaginal discharge with or without other symptoms, attending any of these
clinics, were enrolled into the study after obtaining informed consent. After being routinely
examined by the clinical officer according to the national guidelines using the syndromic
approach, the women were seen by the research doctor. Subjects were interviewed about
their marital, educational, and occupational status using a standardized structured
questionnaire. Sexual, obstetric, and gynecological histories were taken and details of the
current genital tract complaint were noted. Each patient received a full gynecological
investigation including speculum examination and bimanual palpation. Endocervical swabs
were taken for N. gonorrhoeae isolation and C. trachomatis polymerase chain reaction (PCR)
and vaginal swabs for wet preparation, pH testing, and potassium hydroxide testing (sniff
test). The color of the discharge was noted. After pretest counseling a 10 ml sample of
venous blood was drawn for syphilis serology and HIV-1 testing.
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 58

Treat for
cervicitis
Laboratory Procedures
Figure 1: Algorithm for vaginal discharge used as the national policy in Kenya.
Wet mounts were analyzed directly at STC and from the PHC clinics they were transported in
a drop of saline to the laboratory of the department of medical microbiology, University of
Nairobi. The wet mounts were examined for the presence of motile Trichomonas vaginalis
and of yeast cells indicative of Candida albicans. In the laboratory, vaginal smears were heat
fixed, Gram stained, and examined for the presence of clue cells, indicative of bacterial
vaginosis, and for yeast. The diagnosis of C. albicans was made by wet prep and Gram
stain. Bacterial vaginosis was defined by the presence of at least three of the following
criteria: (1) vaginal fluid pH >4.5; (2) release of a fishy amine odor from vaginal fluid mixed
with 10% potassium hydroxide; (3) presence of clue cells; and (4) vaginal discharge (Hillier
1990, Easmon 1992). Cervical swabs for N. gonorrhoeae isolation were inoculated directly
onto Thayer-Martin medium and incubated in a candle extinction jar at 33-35˚ C for 24-48
hours. Cervical swabs for C. trachomatis PCR were processed in the laboratory (Abbott).
Venous blood samples were tested for syphilis using the rapid plasma reagin test (RPR test,
Becton Dickenson) and for HIV-1 using ELISA Detect (Biotech) and Recombigent
(Cambridge).
Data analysis and definitions
Treat for vaginitis
Improvement after 7 days
Patient complains of
vaginal discharge
Lower abdominal pain present
No Yes
Follow PID flow chart
Tenderness on bimanual
examination
No Yes
Yes
No
Discharge
patient
Treat for
cervicitis
Treat for
vaginitis
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 59

Data were entered and analyzed in SPSS for Windows (SPSS, Chicago, IL, USA). In
univariate analysis, the odds ratio (OR) and 95% confidence intervals (CI) were used for the
measurement of associations. Comparisons were made using Pearson's χ 2 and Fisher's
exact tests. Means were compared using the Student's t test. Stepwise logistic regression
was used to take into account risk factors, symptoms, and signs related to the presence of
gonococcal and chlamydial infection.
The principal outcome was the presence of cervical infections. Cervical infections were
defined as the presence of either gonococcal or chlamydial infections or both. Different
algorithms were applied to the study population based on personal data from the interview
and physical examination. We used the results of the analysis of association between risk,
symptoms, and signs to create six different algorithms. Algorithm A was the flow chart as in
use in Kenya (Fig 1). Algorithm B is the Kenyan algorithm but without the abdominal
examination. Algorithm C includes a risk assessment of the patient. This is based on the
WHO algorithm with risk score, but without presence of symptoms in the partner. Also, the
risk assessment we used was simplified and considers the risk score positive if any of the
risk factors are present whereas the WHO algorithm requires at least two risk factors.
Algorithm D is based on the presence of yellow or bloody vaginal discharge. Algorithm E
included presence of endocervical mucopus. And algorithm F combined the risk score and
the presence of yellow or bloody vaginal discharge. The microbiological results were used as
gold standard for cervical infections. The validity of the algorithms was evaluated by
calculating their sensitivity, specificity, and positive predictive value compared with the gold
standard.
We determined the correct treatment rate and the cost for each algorithm. Details on
definitions and calculations are not presented here but can be obtained from the authors.
Results
Socio-demographic Factors and Prevalence of STDs
A total of 621 women with complaints of vaginal discharge were enrolled in the study. In
STC, located in downtown Nairobi, 214 women were recruited and 255 and 152 were
recruited in the two health centers, both located in the suburbs of Nairobi. The women
attending the STD clinic were more often single, employed, and had more sex partners.
Three hundred and thirty four (54%) of the women were pregnant at the time of the interview.
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 60

The demographic and social characteristics of pregnant women compared with non-pregnant
women are shown in Table 1.
Significantly, there were more single non-pregnant women and they had had more sex
partners in the past 3 months. Pregnant women on the other hand were younger.
Table 1: Demographic characteristics of 287 non-pregnant and 334 pregnant women with vaginal discharge in Nairobi, Kenya.
Non-pregnant (n=287) Pregnant (n=334)
n or mean
(% or range)
n or mean
(% or range)
OR (95% CI) P
Mean age 25 (15-52) 23 (16-41) < 0.001
Younger than 20 years 55 (19) 72 (22) 0.9 (0.7-1.2) 0.3
Single 114 (40) 69 (21) 1.9 (1.5-2.5) < 0.001
Housewife/unemployed 146 (51) 220 (66) 0.5 (0.4-0.8) < 0.001
Schooling up to primary 163 (57) 193 (58) 1.0 (0.7-1.3) 0.06
> 1 partner past 3 months 32 (11) 11 (3) 3.4 (1.7-6.6) < 0.001
New partner past 3 months 38 (13) 14 (4) 3.2 (1.7-5.7) < 0.001
Mean age at sexual debut 18 (10-30) 18 (12-30) 0.7
Ever used condoms 128 (45) 112 (34) 1.3 (1.1-1.6) 0.002
Mean number pregnancies 2.0 (0-11) 1.3 (0-6) 0.002
The overall prevalence of reproductive tract infections (RTI) is shown in Table 2. The
prevalence of N. gonorrhoeae and C. trachomatis was 7% and 9% respectively and the
overall cervical infection prevalence was 16%. The majority of women had vaginal infections
caused by C. albicans (51%), T. vaginalis (23%), and bacterial vaginosis (9%). The
prevalence of HIV was 22% and 7% of the women were RPR positive.
Table 2: Prevalence of RTI and HIV among 287 non-pregnant and 334 pregnant women with vaginal discharge.
Total
(n=621)
number (%)
Non-pregnant
women (n=287)
number (%)
Pregnant women
(n=334)
number (%)
Odds Ratio
(95% CI) P
Cervical or vaginal pathogens
N. gonorrhoeae (NG) 44 (7) 34 (12) 10 (3) 4.0 (2.0-7.9)

One of the peripheral health centers had the highest prevalence for almost all conditions
whereas in the STD referral clinic C. albicans was most prevalent. The prevalence of the RTI
for pregnant and non-pregnant women was compared (Table 2). Non-pregnant women had
significantly more N. gonorrhoeae, genital ulcers, and HIV-1 infection. Pregnant women had
more candidiasis.
Variables Related to Gonococcal and Chlamydial Infections
The univariate relation between NG/CT infection and risk factors, symptoms, and clinical
signs for pregnant and non-pregnant women is shown in Table 3. Several risk markers were
associated with NG/CT infection in non-pregnant women, whereas being younger than 20
years old is only associated with NG/CT infection in pregnant women. Having more than one
sex partner is a rare event among pregnant women.
Of the symptoms, only fever is consistently associated with NG/CT infection for both groups
of women although its prevalence is low. Some symptoms on the other hand are more
indicative of not having NG/CT infection. Signs associated with NG/CT infection among nonpregnant
women were yellow or bloody vaginal discharge, presence of purulent or bloody
endocervical mucopus, cervical erythema, and cervical friability. None of these signs was
associated with NG/CT infection in the group of pregnant women.
None of the variables, however, was both sensitive (>60%) and specific (>60%) enough for
the presence of cervical infection. Multivariate analysis including all variables significantly
associated with NG/CT infection (P

Table 3: Association between risk factors, symptoms and signs and gonococcal (NG) or chlamydial (CT) infection among 287 nonpregnant
women and 334 pregnant women, consulting for vaginal discharge*.
Non-pregnant women (n=287) Pregnant women (n=334)
RISK DETERMINANTS
SYMPTOMS
SIGNS
NG/CT -
NG/CT
NG/CT -
NG /CT +ve ve Odds
+ve ve Odds
(n=53) (n=234) ratio
(n=44) (n=290) ratio
n (%) n (%) (95% CI) P n (%) n (%) (95% CI)
Single 27 (51) 87 (37) 1.7 0.05 10 (23) 59 (20) 1.2
(1.0-3.2)
(0.5-2.5)
Age < 20 15 (28) 40 (17) 1.9
(1.0-3.8)
Condom use 22 (42) 106 (46) 0.8
(0.5-1.5)
>1 sex partner 14 (26) 18 (8) 4.3
(2.0-9.4)
1 or more new partners 12 (23) 26 (11) 2.3
(1.1-5.0)
Pruritus 23 (43) 161 (69) 0.3
(0.2-0.6)
Sore vulva 9 (17) 74 (32) 0.4
(0.2-0.9)
Dysuria 15 (29) 57 (25) 1.2
(0.6-2.5)
Dyspareunia 13 (25) 74 (33) 0.7
(0.3-1.4)
Abdominal pain 28 (53) 112 (48) 1.2
(0.7-2.2)
Fever 9 (17) 15 (7) 3.0
(1.2-7.2)
Scratch marks 12 (23) 82 (35) 0.5
(0.3-1.1)
Vulvitis 8 (15) 53 (23) 0.6
(0.3-1.4)
Malodorous discharge 12 (23) 40 (17) 1.4
(0.7-3.0)
Moderate/copious discharge 40 (77) 144 (64) 1.9
(0.9-3.8)
Yellow/bloody discharge 28 (55) 66 (29) 3.0
(1.6-5.5)
Purulent or bloody endocervical
mucopus
23 (43) 55 (24) 2.5
(1.3-4.6)
Cervix erythema 20 (38) 47 (20) 2.4
(1.3-4.5)
Cervix friability 23 (43) 61 (27) 2.1
(1.1-4.0)
Painful adnexa 19 (36) 56 (24) 1.8
(0.9-3.4)
Motion tenderness 12 (23) 33 (14) 1.8
(0.9-3.7)
0.05 20 (46) 52 (18) 3.8
(2.0-7.5)
0.3 12 (27) 100 (35) 0.7
(0.4-1.4)
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 63

Evaluation of Algorithms
The results of the validation of the different algorithms for pregnant and non-pregnant women
are presented in Table 4.
Table 4: Sensitivity, specificity and positive predictive value (PPV) for gonococcal and chlamydial infections for different
diagnostic algorithms for vaginal discharge, among 621 women, and comparing pregnant with non-pregnant women.
Total (n=621) Non-pregnant (n=287) Pregnant (n=334)
Sensitivity Specificity PPV Sensitivity Specificity PPV Sensitivity Specificity PPV
Algorithm (%)
(%) (%) (%)
(%) (%) (%)
(%) (%)
A 42 63 18 47 57 20 36 69 15
B 50 58 18 53 52 20 45 63 16
C 59 61 22 66 54 24 50 67 18
D 45 68 21 55 71 30 33 66 13
E 42 73 22 43 76 29 41 70 17
F 78 42 20 85 38 24 70 46 16
All the above algorithms require the presence of vaginal discharge and:
A: Algorithm as in use in Kenya (complaint of abdominal pain and abdominal tenderness on examination);
B : Modified Kenyan algorithm (complaint of abdominal discharge and no examination performed);
C : WHO algorithm (risk score except for symptoms in partner);
D: Yellow or bloody discharge visible on examination of vagina;
E: Endocervical mucopus on speculum examination;
F: Risk score positive and/or yellow or bloody vaginal discharge visible.
Applying the Kenyan national policy algorithm resulted in a sensitivity of 42% and a
specificity of 63% for gonococcal or chlamydial cervicitis, with a PPV of 18% (algorithm A).
The Kenyan algorithm but without an abdominal examination, hence relying on the complaint
of abdominal pain alone (B), resulted in a higher sensitivity and somewhat lower specificity.
The algorithm with risk score C gave a sensitivity of 59% and a specificity of 61% with a PPV
of 22%. Algorithm D relies on the inspection of the color of the vaginal discharge and does
not necessarily require a speculum examination. Algorithm E requires inspection of the cervix
by speculum examination. Both algorithms D and E result in a sensitivity and a specificity
similar to algorithm A but with marginal higher PPV. Algorithm F, a combination of risk score
and presence of yellow or bloody vaginal discharge, resulted in the highest sensitivity but
had low specificity. Among non-pregnant women, all algorithms produced a higher sensitivity.
Among pregnant women, however, the different algorithms had a lower sensitivity. The
overall correct treatment rate of the different flow charts varied and was in general higher
among non-pregnant women, except for flow charts D and E. The cost per case varied from
US$ 0.5 to US$ 0.9. The cost per true cervical infection treated varied from US$ 6.7 for the
risk score algorithm to US$ 8.3 for the Kenyan algorithm. Algorithm F identified more true
cervical infections than the other flow charts but the overall cost was substantially higher
(table not presented here).
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 64

Discussion
The most common pathogens found among women complaining of vaginal discharge in this
study were Candida albicans and Trichomonas vaginalis. This is not surprising and has been
shown in similar studies among women presenting with vaginal discharge in other African
settings (Costello 1998, Alary 1998, Mayaud 1998). The prevalence of bacterial vaginosis in
our study was low compared with other studies in the region (Wawer 1998, Thomas 1996)
and is probably an underestimation. Indeed, we used the clinical diagnosis for bacterial
vaginosis, which is subject to interpretation by the clinician. Furthermore, the identification of
clue cells depends on the level of training of the person performing the microscopic
examination. Also, the quality of the Gram stain might have influenced the results. The
combination of these factors might have resulted in an underestimated prevalence.
Overall, no cervical or vaginal pathogen or condition was found in 28% of the women. Again,
this is similar to the findings in Mwanza where 23% of the women at the STD clinic and 27%
at the antenatal clinic had no demonstrable pathogen (Mayaud 1998). We found a
prevalence of cervical infections of 16%. Non-pregnant women had significantly more
gonococcal infection, genital ulcers, and HIV infection but pregnant women had more
candidiasis. The high proportion of pregnant women with an STD is a cause for concern.
Gonococcal and chlamydial infections, as well as syphilis and HIV, are associated with
adverse obstetric outcome (Ryder 1991, Temmerman 1992, Leroy 1998, Temmerman 1990).
The results of this study show that there is an urgent need for public health measures to
improve STD control during pregnancy.
The prevalence of the different RTI varied between the different clinics under study.
Surprisingly, the prevalence rates for most RTI were higher in one of the PHC clinics than in
the STD referral clinic. Although the two PHC clinics where enrolment took place are located
very close to each other, the prevalences varied here also, indicating that the populations
attending the different clinics are quite different.
Lower socio-economic conditions seem to be the basis for higher STI prevalences. Health
staff at all levels of medical facilities should become aware that all women are to be
considered at high risk, and that their treatment inclusive of health education and counseling
has to be stressed. The stereotype image that only patients attending the STD clinic are at
high risk has to be revised urgently so as to eliminate the stigma still attached to STDs and
hence to make STD control more effective.
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 65

In this study of symptomatic women, the association of the classic clinical symptoms and
signs with the presence of gonococcal and/or chlamydial infection was quite different among
pregnant and non-pregnant women. Several demographic risk determinants and signs were
predictive of cervical infection among non-pregnant women while hardly any were among
pregnant women. As a result, the various flow charts that were tested performed
systematically better in the group of non-pregnant women. The algorithm in use in Kenya as
national policy had a sensitivity of 42% and a specificity of 63% for the detection of N.
gonorrhoeae or C. trachomatis and thus failed to discriminate between infected and
uninfected women. This algorithm relies on bimanual examination of the patients, and hence
requires an examination table and gloves, items often not available in the health centers. As
a result, the algorithm is, in practice, often applied without performing this examination.
Furthermore, the bimanual examination is subject to interpretation and depends on the
experience of the person performing it. In our study the physical examinations were done by
medical doctors, and they reported almost all women in the study to have abdominal
tenderness. In practice in the health centers, the examinations are performed by nurses who
are less well trained; hence we can assume that the results would be worse. We therefore
also tested the Kenyan algorithm but without performing the bimanual examination (algorithm
B). The results are similar and in fact have a higher sensitivity.
The risk score used to discriminate cervical infection, as promoted by the WHO, has been
evaluated in several African settings (Mayaud 1998, Thomas 1996, Vuylsteke 1993, Diallo
1998, Mayaud 1995, Germain 1997). We used a simplified risk score without symptoms in
the partner as this information is sensitive and difficult to obtain. Personal communication
with health care workers and data from our other studies (unpublished) suggested that
symptoms in the male partner are seldom known by the woman and are hence unreliable.
This was confirmed in a study by Thomas et al. who found that only 2% of mostly married
women reported partners having symptoms. Hence, including this risk factor would only
increase the specificity but not the sensitivity of the flow chart. The algorithm we tested (C)
was based on presence of any of the risks: being single, being less than 20 years old, having
had multiple or a new partner in the past 3 months. An earlier study in Nairobi also reported
being single and having multiple sex partners to be associated with gonococcal infection in
pregnant women (Costello 1994). We considered that for a risk score to be applicable it has
to be simple to use. The health care worker in a busy health center has no time and patience
to apply scoring systems with weighted risks or calculations of scores. In our study a risk
score based on presence of at least two of the risks would have resulted in a sensitivity of
only 30%.
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 66

Our risk score algorithm (C) resulted in a sensitivity of 66% with PPV of 24% among nonpregnant
women. The results among pregnant women were less good and comparable with
what Mayaud et al. had found in Mwanza. Using this flow chart, the rate of overtreatment in
non-pregnant women was 46% while it was 33% in pregnant women.
We included several signs in the flow chart. Among pregnant women, however, none of
these signs was associated with cervical infection. This is contrary to findings from Thomas
et al. who found that among asymptomatic women in Nairobi, cervical friability was
associated with cervical infection. An earlier study in Nairobi among pregnant women had
identified friability and endocervical mucopus as predictors of cervical infection (Braddick
1990). And in another study in Nairobi, Temmerman et al. reported an association between
gonococcal infection post partum and cervical mucopus. We tested algorithm D based on
observation of the color of vaginal discharge. This algorithm does not require the use of a
speculum. Algorithm E was based on observation of the cervix hence requiring speculum
examination that is seldom possible in health centers in Kenya. Both resulted in too low
sensitivity to be of value. Algorithm F combined the risk assessment and the inspection of the
vaginal discharge. In both groups of women high sensitivity was reached but with low
specificity. Several other combinations of risks, symptoms, and signs were tested but all
performed worse.
While none of the tested algorithms reached acceptable levels of sensitivity and specificity,
the algorithm with risk score performed somewhat better than the algorithm actually in use in
Kenya and would identify more true cervical infections although at a higher overall cost.
Introducing the risk assessment among non-pregnant women could be an option. It is,
however, doubtful that the introduction of the risk score into the existing algorithm, which
would imply printing of new charts and retraining of health staff, is worthwhile. Among
pregnant women this flow chart would fail to identify cervical infections. Treatment of
pregnant women with vaginal discharge might be considered for both vaginal and cervical
infections on the first visit. Further operational research is needed to assess the rate of follow
up and return visits among these women. We can conclude that simple, cheap, and reliable
tests for the diagnosis of cervical infection in women are still urgently needed.
DIAGNOSIS AND MANAGEMENT OF SEXUALLY TRANSMITTED INFECTIONS 67

C HAPTER 6
PREVENTION OF SEXUALLY TRANSMITTED
6.1. Syphilis Control During Pregnancy:
INFECTIONS INCLUDING HIV
Effectiveness and Sustainability of a Decentralized Program.
Summary
Published in the American Journal of Public Health 2001;
91:705-7 by Fonck K, Claeys P, Bashir F, Bwayo J, Fransen L
and Temmerman M.
This study sought to assess the performance, effectiveness, and costs of a decentralized
antenatal syphilis-screening program in Nairobi, Kenya. Health clinic data, quality control
data, and costs were analyzed. The rapid plasma regain (RPR) seroprevalence was 3.4%. In
terms of screening, treatment, and partner notification, the program’s performance was
adequate. The program’s effectiveness was problematic because of false-negative and falsepositive
RPR results. The cost per averted case was calculated to be US$ 95 to US$ 112.
The sustainability of this labor-intensive program is threatened by costs and logistic
constraints. Alternative strategies, such as the mass epidemiologic treatment of pregnant
women in high-prevalence areas, should be considered.
Introduction
Pregnant women who are seroreactive to syphilis are at an increased risk for spontaneous
abortion, stillbirth, prematurity, and perinatal death (Schultz 1990, Watts 1984, Temmerman
1992, Berkowitz 1993, How 1994). In Africa, the reported prevalence of syphilis in pregnancy
ranges from 3.6% to 19% (Meda 1997, Grosskurth 1995, Wawer 1998, Mayaud 1997,
Wilkinson 1997, Newell 1993, Temmerman 1992). Although syphilis screening in pregnancy
is a national health policy in most African countries, few screening programs achieve
implementation, mainly owing to financial and logistical constraints.
Chapter 6 68

One of the priorities of the Kenyan Sexually Transmitted Diseases Control Program, set up in
1987 (Forsythe 1996), was to control syphilis during pregnancy. In 1989, an evaluation of the
centralized screening program in Nairobi revealed low effectiveness (Temmerman 1993).
Subsequently, in 1992, a decentralized program was put in place with support from the
European Commission. This program has since been extended to other locations in Kenya
(Jenniskens 1995).
The purpose of this study was to assess the effectiveness of this 6-year-old urban
decentralized program and to formulate recommendations for future implementation.
Methods
The main strategy of the syphilis program in Nairobi was the implementation of a
decentralized, clinic-based model for the on-site diagnosis and treatment of syphilis
seroreactive pregnant women and their partners. Components of the project included:
(1) laboratory support, (2) supplies and drugs, (3) training nurses in rapid plasma reagin
(RPR) testing and treating seroreactive women with a single dose of 2.4 million international
units of benzathin penicillin administered intramuscularly, (4) counseling, (5) partner
notification, and (6) supervision and monitoring. The Macro-Value RPR card test (Becton
Dickinson, Cockeysville, Md) was used for RPR testing at the peripheral level and for quality
control at the central sexually transmitted disease (STD) referral clinic. The quality control
consisted of a blind analysis of all RPR-positive sera and of a random selection of 1 in 10
negative samples.
To analyze the program, we monitored data from the 10 primary health care clinics, quality
control data from the referral laboratory, and information on costs.
Results
Between July 1997 and June 1998, 96% of all pregnant women (n=27,377) attending the 10
clinics were screened for syphilis. Main reasons for failure to screen were high workload and
lack of electricity, although manual RPR cards were available. Overall, 928 women (3.4%)
were RPR positive, and 91 % of these women were promptly treated (Table 1). The reasons
for non-treatment included absence of the person in charge, unwillingness of the woman to
wait for test results, and the policy of some of the clinics to treat the woman simultaneously
with the partner.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 69

Table 1: Syphilis Rapid Plasma Reagin (RPR) seroreactivity and treatment rates at 10 Mother
and Child Health (MCH) Clinics in Nairobi, Kenya.
Women screened RPR+ Women Treated Partners Treated
MCH clinic n % n % n % n %
Baba Dogo 2863 97 160 6 153 96 64 42
Dandora 4064 97 183 5 173 95 112 65
Eastleigh 1453 86 51 3 50 98 27 54
Kariobangi 2647 85 30 1 23 77 19 83
Langata 4662 100 168 4 149 89 74 50
Mathare 2793 100 78 3 78 100 40 51
Ngong Road 2749 100 75 3 58 77 44 76
Riruta 1651 90 41 2 33 80 11 11
Umoja 2458 100 90 4 80 89 41 41
Westlands 1947 100 52 3 48 92 20 20
Total 27377 96 928 3 845 91 452 452
Overall, 53% of the partners were treated. Partner notification was hampered by the quality
of counseling, the casual nature of the partnerships, and a woman's risk of physical abuse
when she informed her partner of the STD; hence, women might be reluctant to do so
despite the counseling's focus on the well-being of the unborn child.
When we performed quality control, only 69% of the RPR-positive results and 97% of the
RPR-negative results were confirmed. The relatively high false-positive rate may be
explained by technical constraints, high workload inadequate training, and slackening of the
supervision. There is also anecdotal evidence that some centers report false-positive results
on purpose in order to obtain benzathin penicillin for other uses.
The costs of the program include testing, Table 2: Cost in Kenyan Shilling of the syphilis screening program
treatment, quality control, yearly
during a one-year period, (1 US Dollar = 60 Kenyan Shillings).
No of Units Cost / Unit Total Cost
refresher training, and supervision. The RPR kits 200 2000 400000
capital costs (car, rotators, and Vacutainers and needles 27377 2 54754
centrifuges) were discounted over 5
Rotators and centrifuges* 10 20000 200000
Drugs 1420 25 355000
years. The total cost of the syphilis Staff salaries** 372000
screening program for 1 year was US$ Refresher training 1 100000 100000
30,996 (1 US dollar= 60 Kenyan
Car*
Transport cost
1 200000
82000
shillings) (Table 2).
Total
1859754
The total cost of testing and treatment Cost per pregnant woman
68
(RPR kits, Vacutainer collection tubes
* Capital cost discounted over 5 years.
and needles, rotators and centrifuges,
** Includes only salaries for staff involved in supervision and quality
control, does not include clinic staff salaries.
clinic supplies, and drugs) was US$ 18,429. Of the 27,377 women screened, 928 were RPR
reactive and 845 were treated, resulting in a cost of US$ 37 per case treated. Positive results
were confirmed in only 583 (69%) of the women treated. If congenital syphilis occurs in one
third of babies born to syphilitic mothers and stillbirth or prematurity in another third, the
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 70

program has prevented the harmful effects of syphilis on the unborn child in 388 women,
given that treatment is 100% effective. This would result in a cost of US$ 56 per averted
case. The real effectiveness can be estimated at 50% to 70% because of late treatment or
reinfection. Hence, the cost per case averted will vary between US$ 95 and US$ 112.
Discussion
Syphilis seroprevalence among pregnant women in Nairobi has decreased significantly in
recent years. The RPR seroprevalence was 7.2% in 1993, 4.5% in 1996, and 3.4% in 1998
(Temmerman 1999). This decline might be the result of different intervention programs and
improved health care services, including the syphilis control program, which may have had
an independent effect. Hence, the program may also have had an indirect influence on the
spread of HIV in this population.
Overall, the clinics performed well in screening and treating RPR-seropositive women and
their partners. The partner notification rate of 53% reported here is twice as high as for nonpregnant
patients with an STD other than HIV who attended the same Clinic (Gichangi
2000). Concern about the unborn baby seems to motivate future fathers to get treatment
(Pido 1993, Njeru 1995) and may help women to inform their partners despite the threat of
possible physical abuse (Temmerman 1995).
The effectiveness of the treatment may be limited because women in Kenya tend to come
late for prenatal care (Temmerman 1993, Pido 1993). It is also hampered by weak screening
performance in field conditions. In 1993, only 17% of RPR-positive results and 1.2% of RPRnegative
results were found to be false, but monitoring and external supervision were very
intensive in the early years (Jenniskens 1995). The observed reduction in the quality of the
test over a 6-year period raises questions about the viability of the programs. In addition,
false-positive reactions might be higher in the presence of HIV (Nandwani 1995, Joyanes
1998, Fonck 2000).
Cost-effectiveness estimates of syphilis screening during pregnancy vary widely, depending
on the underlying assumptions (Hira 1991, Stray-Pedersen 1980, Abyad 1995, McDermott
1993, Schmid 1996). According to a World Bank study, the prevention of congenital syphilis
by routine screening and treatment is one of the most cost-effective interventions for
improving child health (World Bank 1993). Our findings underscore this statement, but they
also stress technical, logistic, and financial drawbacks that have been pointed out by others
(Temmerman 1993, Jenniskens 1995, Rutgers 1993).
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 71

One should keep these problems in mind when planning HIV mother-to-child transmission
programs, an area where testing and antiretroviral treatment have many more implications
than with syphilis testing.
The question remains whether case detection and treatment is the best option. An alternative
strategy could be mass treatment with penicillin in pregnancy. With a cost of US$ 1.50 per
dose and an estimated acceptability of 75%, the total cost for such a program would be US$
30,799, similar to the cost of US$ 30,966 for the current program in Nairobi. With this
strategy, 1075 true-positive cases (75%) would be treated compared with 583 now. If the
treatment is effective in 50% of the cases, the cost per averted case of congenital syphilis
would be US$ 86, a reduction of 23%. The systematic administration of antibiotics during
pregnancy may also have a positive impact on pregnancy outcome (Temmerman 1995,
Gichangi 1997). Similarly, routine treatment would reduce the negative aspects of contact
tracing. Concerns of mass treatment, however, are the potentially increased resistance
patterns and the massive use of needles and syringes.
Syphilis control during pregnancy is an effective intervention, especially in high-prevalence
areas, but it requires careful monitoring, quality control, and close supervision.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 72

6.2. Partner notification of pregnant women infected with syphilis in
Nairobi, Kenya.
Summary
Published in the International Journal of STD & AIDS 2000;
11:257-61 by Gichangi P, Fonck K, Sekande-Kigondu C,
Ndinya-Achola
Temmerman M.
J, Bwayo J, Kiragu D, Claeys P and
We examined partner notification among syphilitic pregnant women in Nairobi. At delivery,
377 women were found to be rapid plasma reagin (RPR) reactive. Data were available for
94% of the partners of women who were tested during pregnancy; over 67% of the partners
had received syphilis treatment while 23% had not sought treatment mainly because they felt
healthy. Six per cent of the women had not informed their partners as they feared blame
and/or violence. Adverse pregnancy outcome was related to lack of partner treatment during
pregnancy (7% versus 19%, odds ratio (OR) 3.0, 95% confidence interval (CI) 0.9-10.0). Our
data suggest that messages focusing on the health of the unborn child have a positive effect
on partner notification and innovative and locally adapted strategies for partner notification
need more attention.
Introduction
Partner notification contributes to the control of sexually transmitted diseases (STDs) by
identifying and treating previously undiagnosed infection the sexual contacts of known cases,
thereby preventing re-infection of the index case or transmission to others (Potterat 1991).
The effectiveness of the strategy depends upon the ability of the patient or the health advisor
to locate contacts and ensure treatment (Bell 1998). In resource-poor settings with limited
public health infrastructures, patient referral may be the most feasible method.
The proportion of partners notified and treated has been used as a measure of program
success but depends on various factors such as prevalence and symptomatology of the
disease, sexual behavior of the index patient and motivating factors (Gunn 1997, Andrus
1990, Jenniskens 1995). In Zimbabwe, partner referral increased from 19% to 42% through
outreach workers (Winfield 1985).
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 73

The use of mailed reminders did not increase the rate of partner referral in Nigeria, as the
addresses given by the index patients were mostly incorrect (Asuzu 1984, Asuzu 1990).
In industrialized countries, prompt treatment of infected individuals and their partners is seen
as a strategy to eradicate syphilis (Gunn 1997). In the USA, partner notification played an
important role in reduction of incidence of syphilis from 72 cases per 100,000 in 1943 to 4
cases per 100,000 in 1956 (Brown 1970, Brandt 1988).
In Kenya, syphilis control in pregnancy is national health policy and is part of the strategy for
antenatal care in both the government and private sectors. Partner notification and
counseling however were not considered part of the antenatal syphilis control by Nairobi City
Council staff providing antenatal care until after an evaluation of syphilis control activities in
1993 (Temmerman 1993). Since then, partner notification has become an integral part of
syphilis control in pregnancy. All women attending antenatal care are tested for syphilis using
the RPR card test. Women who are RPR reactive are treated and encouraged to bring their
partner(s) for treatment. As a result, partner notification reached 50% in 1993 (Jenniskens
1995).
In this paper, we report the determinants and effects of partner notification and treatment in
pregnant women in Nairobi, Kenya.
Subjects and methods
As part of a study on the effectiveness of syphilis screening and treatment during pregnancy,
we interviewed women at the main maternity hospital in Nairobi after the delivery. After giving
informed consent, they were tested for syphilis and HIV-1. Women who were seroreactive
were treated and requested to bring their partners for treatment. They were given a verbal
message and partner notification coupons. The importance of treatment of partner(s) in
relation to adverse events for future pregnancies and re-infection was emphasized.
A structured questionnaire was administered to all women who were RPR positive during
pregnancy and/or at delivery. The questionnaire included demographic, medical and
obstetrical factors.
Partners who presented to the study clinic were treated, counseled and were given the
option to be tested for syphilis or not. Partners who consented to HIV testing were given HIV
pre-test counseling.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 74

Syphilis tests were done using the RPR card test (Becton Dickinson, Maryland, USA) and
positive samples were confirmed with Treponema pallidum haemagglutination test (TPHA,
Randox Laboratories, UK). Serum HIV-1 antibodies were detected by enzyme-linked
immunosorbent assay (ELISA) (Biochem immuno systems kit, Montreal, Canada) and
positive samples were confirmed using double ELISA test (Biotech, Cambridge, Ireland).
Data analysis was done using SPSS version 7.5. Pearson's Chi-square with Yates' correction
was used to compare proportions and the t-test to compare means. ORs and their 95% CIs
were used to measure strength of association.
The study was approved by the Ethical Committee of the University of Nairobi.
Results
From April 1997 to April 1998, an unselected group of 12,414 women who gave birth in the
Pumwani Maternity Hospital (PMH) were screened for syphilis. Of them, 377 (3%) were RPR
reactive. Of those 377 women, 117 (31%) had already been identified RPR positive earlier
on during pregnancy while 155 (41%) had been RPR negative and 105 (28%) had not been
tested during pregnancy.
Postpartum, partner notification information was provided to 357 of the 377 RPR-positive
women and 127 (36%) of their partners came for treatment. Forty-two (33%) of those
partners had already been treated during pregnancy and of them 12 (29%) were still RPR
positive at the time of delivery. Twenty-two men (17%) declined to be tested for syphilis. Of
the 105 men tested, 33 (31%) were RPR positive. Fourteen men had a positive RPR test but
negative TPHA test, suggesting a new early infection or a false positive RPR test.
Table 1 shows the characteristics of women whose partners came for treatment and those
who did not turn up. There was no difference in education, history of alcohol use, parity, bad
obstetrical history, HIV status or number of sex partners in the past year. Women whose
partners did not come for treatment however were younger (P=0.01), more likely to be
unmarried (P

Table 1: Characteristics of women whose partners came and whose partners did not come for treatment.
Characteristic Women whose partners did not
come for treatment
n=230 %
Women whose partners came for
treatment
n=127 %
P- value
Age in years (mean + SD) 23 + 4.7 24 + 4.5 0.02
< 20 yrs old 87/230 (38) 30/123 (24) 0.01
Single 57/228 (25) 6/123 (5) 0.000
RPR + during pregnancy 67/161 (42) 49/100 (49) 0.2
Sex in exchange for money 32/230 (14) 5/123 (4) 0.004
Sex with person with STD 16/227 (7) 8/120 (7) 0.9
Ever had STD 25/230 (11) 22/123 (18) 0.06
Unfaithful spouse 61/169 (36) 25/101 (25) 0.05
History of abortion 39/230 (17) 21/124 (17) 1.0
History of LBW 13/100 (13) 3/59 (5) 0.2
Nullipari 102/230 (44) 50/124 (40) 0.5
Primigravida 85/230 (37) 43/124 (30) 0.7
Inadequate antenatal care 101/229 (44) 37/123 (30) 0.01
HIV + at delivery. 62/229 (27) 24/127 (19) 0.07
< 8 yrs of education. 161/229 (70) 77/123 (63) 0.1
Adverse obstetrical outcome 54/228 (24) 13/122 (11) 0.004
>1 sex partner in past year 92/230 (40) 40/123 (36) 0.2
Antenatal care in intervention clinic 90/230 (39) 58/123 (47) 0.1
Alcohol use 32/229 (14) 9/123 (7) 0.06
Treated for syphilis during pregnancy 63/67 (94) 49/49 (100) 0.08
Table 2 compares behavior
Table 2: Characteristics of women and their partners who came for treatment.
characteristics of women and
their partners who came for
Characteristic
Women
n=127 (%)
Partners
n=127 (%)
P
treatment. Men were more likely Age years (mean + SD) 24.1 + 4.5 30 + 6
to have had sex with commercial
Single 6/123 (5) 6/127 (5) 1.0
Sex in exchange for money 5/123 (4) 19/127 (15) 0.003
sex workers (P=0.003), to have Sex with person with STD 8/120 (7) 29/127 (23) 0.000
had sex with a person with an Ever had STD 22/123 (18) 78/126 (62) 0.000
STD (P

Table 3. HIV status of rapid plasma reagin seropositive
women and their partners
Partner HIV + Partner HIV - Total
HIV + women 7 (8%) 13 (14%) 20
HIV - women 4 (4%) 68 (74%) 72
Total 11 81 92
Of the 12,414 women tested at delivery, 7644
(62%) had already been screened for syphilis
during pregnancy. Of them, 228 (3%) were
reported RPR reactive and almost all of them
(98%) had been treated during pregnancy.
Information on partner notification and treatment was available for 215 (94%) of the RPRpositive
women. About 67% of the partners had received treatment while their spouses were
pregnant. RPR-positive women whose partners had been treated during pregnancy had less
adverse obstetrical outcome, defined as low birth weight and stillbirth, than women whose
partners were not treated (7% vs. 19%, OR 3.0, 95% CI 0.9-10.0, P- value 0.09).
The women reported several
reasons for non-treatment of their
Table 4. Reasons for non- treatment of partners by women’s marital status.
partners (Table 4). The main
reason for non-treatment of married
Reason for partner not being treated
Single
n=21 (%)
Married
n=50 (%)
P
partners was not feel sick. Single
Not feeling sick
Not informed
1 (5)
8 (38)
27 (54)
4 (8)
0.000
0.004
women more often did not inform Casual partner 9 (43) 1 (2) 0.000
their partners or could not trace the Unknown treatment status 3 (14) 18 (36) 0.1
casual partner. Over 17% (12/71) of the women had not informed their partners because of
fear of violence or being blamed for the illness. Significantly more partners of married women
were treated during the index pregnancy (135/185 vs. 9/29, P

studies, indicating that couples are more accepting and compliant with partner treatment
recommendations when they are framed in the context of reproductive health (Desormeaux
1996, Jenniskens 1994). In Rwanda, pregnant women were more likely to refer partners than
non-pregnant women with an STD (33% vs. 20%) (Steen 1996). In an earlier study from
Kenya, 50% of partners of pregnant women with syphilis were treated in the same antenatal
clinic as the index patients (Jenniskens 1995). In Haiti, 41% of the partners of pregnant
women with STDs, were treated mainly as a result of index referral (Desormeaux 1996).
The referred partners are, with few exceptions, the regular partners of tile index case.
Others have also shown difficulties of tracing casual contacts, which may play an important
role in continued transmission (Steen 1996, Day 1998). Moreover, none of the women in our
Study referred more than one sexual partner for treatment despite the fact that 33% of RPR
seroreactive women admitted having had more than one sexual partner in the past year.
Women with syphilis are said to be more promiscuous than other STD patients as reported
by Andrus et al. who stated that patients who had syphilis had a larger number of sexual
encounters with persons who subsequently could not be identified unlike patients who had
gonorrhea. Unmarried women were also less likely to refer a sexual partner for treatment.
This suggests that more innovative methods and strategies are needed for reaching casual
partners.
There are several possible areas for improving partner referral. Improved counseling of index
patients is the most straightforward method. Educational messages at the community level
emphasizing the importance of partner treatment might also improve partner referral rates.
More innovative approaches are needed to identify the more epidemiologically important
casual partner. One strategy could be active tracing by specially trained staff (provider
referral) as opposed to notification by the index patient (patient referral) (Potterat 1991).
However, the logistics and cost of active tracing pose serious limitations to this approach,
and this is not only in resource-poor settings. In the UK, only 20 % of all sexual contacts of
patients with gonorrhea could be actively traced and subsequently screened (Bell 1998). A
combination of different methods might be helpful in further reducing the syphilis
seroprevalence that has been shown to decrease among pregnant women in Kenya
(Temmerman 1999).
All tracing programs however should take into account possible implications on the harmony
of the family. As shown in other studies, women who inform their spouse of having an STD
often face violence or even break-up of marriages or are blamed for bringing the disease to
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 78

the family (Temmerman 1995). Programs involving partner notification should weigh the
benefits and risks for the women involved.
We have shown a marked difference in pregnancy outcome between women whose partners
were treated during pregnancy and those who were not. Although the figures are small and
the difference does not reach statistical significance, we think that this effect is a real one.
We found an HIV seroprevalence of 22% in RPR-positive women and 12% among the
partners tested. The HIV discordance rate was 18%, which corresponds to what has been
reported in Uganda (Serwadda 1995), Zambia (McKenna 1997) and Rwanda (King 1993).
We found that 35% of the partners of HIV-seropositive women were positive compared with
80% in Rwanda (Allen 1992), 15% in Zaire (Ryder 1990) and 5% in the UK (Johnson 1989).
These differences might be explained by the fact that all women in our study were RPR
positive. They might be at higher risk of acquiring HIV through their high-risk behavior or
through the interaction of the STD with the acquisition of HIV infection (Wasserheit 1992).
In conclusion, innovative strategies to improve partner notification are needed, but should
take into account the potential disastrous consequences of increased violence, divorce or
loss of security. Targeted education could reduce this problem making it possible for the full
benefits of partner notification to be appreciated. Studies to design strategies that can
motivate more partners to go for treatment after delivery are needed.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 79

6.3. A Randomized, Placebo-controlled Trial of Monthly
Azithromycin Prophylaxis to Prevent Sexually Transmitted
Infections and HIV-1 in Kenyan Sex Workers: Study Design and
Baseline Findings.
Summary
Published in the International Journal of STD & AIDS 2000;
11:804-11 by Fonck K, Kaul R, Kimani J, Keli F, MacDonald K,
Ronald A, Plummer F, Kirui P, Bwayo J, Ngugi E, Moses S and
Temmerman M.
Our objectives were to describe the baseline findings of a trial of antibiotic prophylaxis to
prevent sexually transmitted infections (STIs) and HIV-1 in a cohort of Nairobi female sex
workers (FSWs). A questionnaire was administered and a medical examination was
performed. HIV-negative women were randomly assigned to either one gram azithromycin
or placebo monthly. Mean age of the 318 women was 32 years, mean duration of sex work
7 years and mean number of clients was 4 per day. High-risk behavior was frequent: 14%
practiced anal intercourse, 23% sex during menses, and 3% used intravenous drugs. While
20% reported condom use with all clients, 37% never use condoms. However, STI
prevalence was relatively low: HIV-1 27%, bacterial vaginosis 46%, Trichomonas vaginalis
13%, Neisseria gonorrhoeae 8%, Chlamydia trachomatis 7%, syphilis 6% and cervical
intraepithelial neoplasia (CIN) 3%. It appears feasible to access a population of high-risk
FSWs in Nairobi with prevention programs, including a proposed trial of HIV prevention
through STI chemoprophylaxis.
Introduction
Sexually transmitted infections are among the most common diseases worldwide, with major
medical, social and economic consequences. The prevalence of STIs in the general
population in Kenya is unknown, but has been shown to exceed 20% in pregnant women in
Nairobi. The prevalence of N. gonorrhoeae, C. trachomatis, Treponema pallidum and HIV-1
among pregnant women were 7%, 9%, 6% and 15% respectively (Temmerman 1994).
Surveillance data from antenatal clinics in Nairobi in 1996 revealed an HIV-1 prevalence of
16% (Jackson 1999).
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 80

In some parts of Africa, FSWs and their clients are considered to be the group at highest risk
of HW infection and hence constitute an important population for targeted interventions.
Studies among sex workers in Nairobi have shown a HIV prevalence rate of 67% and an
annual HIV incidence of 47% (Plummer 1991).
There is considerable evidence that both ulcerative and non-ulcerative STIs enhance HIV-1
transmission and acquisition (Wasserheit 1992, Sewankambo 1997, Kreiss 1994, Moss
1995, Laga 1993). STI control may therefore offer a means of reducing the spread of the
AIDS epidemic. A randomized community trial in Mwanza, Tanzania demonstrated that
improved management of symptomatic STIs resulted in a 40% reduction in HIV-1 incidence
in the intervention arm (Grosskurth 1995). However, a randomized community trial of
antibiotic prophylaxis every 10 months in Rakai, Uganda, failed to demonstrate any reduction
in HIV-1 incidence, despite a significant reduction in the incidence of T. vaginalis and T.
pallidum infection (Wawer 1999). It has been suggested that this discrepancy may be due to
the maturity of the HIV-1 epidemic in Rakai at the time of the trial, which meant that most
HIV-1 transmission was no longer driven by STIs (Hitchcock 1999). However, rates of N.
gonorrhoeae, C. trachomatis and Haemophilus ducreyi were not significantly reduced by
antibiotic prophylaxis, possibly due to their low frequency in the target population. In addition,
the administration of antibiotic prophylaxis every 10 months may be inadequate, since it may
have little effect on the natural history of STIs such as gonorrhea, which frequently resolve
within weeks, even if untreated and which may quickly be re-acquired after treatment
(Handsfield 1995, Costa 1985).
A more feasible and logical approach in the face of a mature HIV-1 epidemic may be to
target frequent antibiotic prophylaxis towards high-risk 'core' transmitter groups, such as sex
workers, who are at high risk of acquiring both STIs and HIV-1 infection instead of targeting
unselected populations.
Hence we decided to set up a randomized trial of monthly azithromycin prophylaxis to
prevent STIs and HIV-1 infection in a cohort of Kenyan sex workers from the Kibera district of
Nairobi. Azithromycin, a macrolide antibiotic with a favorable side effect profile and a long
tissue half-life, has demonstrated efficacy against the STI pathogens N. gonorrhoeae, C.
trachomatis and H. ducreyi (Ridgway 1996).
The objectives of this study were: (1) to evaluate the impact of regular routine azithromycin
chemoprophylaxis on STI and HIV-1 incidence in this population; (2) to examine potential
adverse effects of monthly azithromycin administration, including effects on antibiotic
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 81

esistance patterns in bystander flora. This paper describes the study design, and the
baseline characteristics of the first 318 sex workers screened in the study.
Methods
The study design was an open, randomized placebo-controlled clinical trial. Half of those
enrolled received monthly azithromycin prophylaxis during 2 years. Only HIV-seronegative
women were enrolled but all women were provided with the appropriate medical and
counseling services and health education. From previous experience in Kenya, the
cumulative risk of HIV-1 seroconversion among the FSWs was estimated to be
approximately 30% over 2 years (Plummer 1991) and it was assumed that monthly
prophylaxis with azithromycin would reduce the incidence to about 15%. With a β error of
0.2 (power of 80%) and a 2-sided I error of 0.05, the required sample size was approximately
120 in each group. A loss to follow-up of about 30% was expected resulting in a required
sample size of about 170 in each group.
The HIV seroprevalence in these women was estimated to be approximately 50% so about
680 women would have to be screened to identify 340 seronegative women.
The FSWs were recruited through a network of peer-leaders. FSWs who were interested in
joining the study presented themselves at the study clinic. At the initial visit, after informed
consent and HIV pre-test counseling, a structured questionnaire on sexual behavior,
reproductive and medical histories was administered. A full physical examination including
speculum examination and laboratory STI and HIV testing was performed.
At the subsequent visit, HIV-1 seronegative women were invited for enrolment into the trial.
More counseling on the nature of the trial, and emphasis on the need for reduction in risk
behavior, was given. An information form was provided to the women who were asked to
return to the clinic if they should decide to join the study.
Block randomization to either azithromycin one gram monthly or placebo took place. The
study medication was given monthly in the clinic under direct observation. At each visit a
urine specimen was obtained for later analysis for asymptomatic STIs (gonorrhea and
chlamydia) using polymerase chain reaction (PCR). Any symptomatic STIs were treated
according to the Kenyan National AIDS/STD Control Program guidelines. At each visit, all
women were extensively counseled about the need for risk reduction.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 82

HIV-1 serology was performed every 3 months. A full medical examination, with STI
screening and treatment was performed every 6 months. Free medical care was offered to all
FSWs who volunteered for screening, regardless of the HIV-1 serostatus. Condoms were
also provided free of charge through the clinic, along with health education on safe sex
methods.
The study was approved by the Ethical Review Committees of the University of Nairobi and
of the University of Manitoba. Informed consent was obtained from all study participants. A
drug safety and monitoring board (DSMB) was established to monitor the study closely.
Vaginal swabs were taken for T. vaginalis culture, pH and Gram stain. Cervical swabs were
obtained for N. gonorrhoeae and C. trachomatis PCR, for N. gonorrhoeae culture, and for a
Pap smear. If a genital ulcer was present, the ulcer base was swabbed for culture and
multiplex PCR (M-PCR). Blood was drawn for HIV-1 and syphilis serology.
Vaginal pH was read using pH indicator strips (Merck, Darmstadt, Germany). Vaginal smears
were heat-fixed and Gram stained. Bacterial vaginosis was defined using a quantitative
scoring system (Nugent 1991). T. vaginalis culture was performed using the In Pouch TM TV
(Biomed Diagnostics, San Jose, California). The Pap smear, taken by cytobrush, was read
using the Bethesda classification (Nat Cancer Instit 1989).
The swab for N. gonorrhoeae was inoculated directly onto Thayer-Martin medium and
incubated at 37ºC, 5% CO2 for 24-48 h. H. ducreyi culture was performed on an activated
charcoal medium (Lockett 1991). PCR testing was performed for C. trachomatis and N.
gonorrhoeae (Amplicor PCR Diagnostics, Roche Diagnostic System, Ontario, Canada). If a
genital ulcer was present, a swab of the ulcer base was used for M-PCR detection of H.
ducreyi, herpes simplex virus and T. pallidum (Roche Molecular Systems, Ontario, Canada).
The rapid plasma reagin test (RPR test, Becton Dickinson, Groot-Bijgaarden, Belgium) was
performed for syphilis serology and positive samples were confirmed by Treponema pallidum
haemagglutination assay (TPHA) (Randox Laboratories, UK). HIV-1 screening ELISA was
performed using the Detect-HIV kit (BioChem ImmunoSystems Inc, Montreal, Canada) and
positive tests were confirmed with the Recombigen HIV-1/HIV-2 EIA (Cambridge Biotech
Corporation, Galway, Ireland).
Data were entered in ACCESS Microsoft, and analysis performed after export into Statistical
Package for Social Sciences (SPSS) for Windows version 8.0. (SPSS Inc. Chicago).
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 83

In univariate analysis, the odds ratios (OR) and 95% confidence intervals (CI) were used for
the measurement of association of proportions. Comparisons were made using Pearson's χ 2 ,
Fisher's exact tests and χ 2 test for trend. Student's t-test was used for comparison of means.
Stepwise logistic regression was used to take into account risk factors and presence of other
STIs for the risk of HIV infection.
Results
Between May and December 1998, 318
women were screened (Table 1). Two women
deferred examination after interview. The
mean age was 32 years. Although none of the
women were married, 60% of them reported
having at least one regular partner. They had
been in prostitution for a mean of 7 years. The
average number of sexual encounters was 4
per day and 16 per week. The average
amount charged per sex act was 118 Kenyan
Shilling (~2 US dollar). Condom use in this
population was variable: while 42% of FSWs
reported using condoms more than half the
time, 37% never used condoms. Most of the
women worked at home, fewer worked in bars
and a minority worked in nightclubs.
About half of the women did not use any form
of contraception. Of the remainder, the
contraceptive method most commonly used
was Norplant (17%) followed by oral
contraceptives (11%). Condoms as a form of
contraception were reported by 11% of the
women. Sixteen women were pregnant.
Vaginal douching, understood as inserting
fluids in the vagina, was practiced by 69% of
the women, and most did so after every
intercourse.
Table 1. Demographic, sexual and high- risk behavior
and medical history of 318 sex workers in Kibera, Nairobi.
n or mean % or range
Age 32 18-57
Resident in Nairobi
Marital status
11 1-50
Never married
131 41
Widowed/divorced/separated 186 59
Age at first sex 16 9-28
Duration of prostitution in years 7 0-34
Partners per day 4 1-10
Partners per week 16 1-70
Charges for sex act (in KSh)
Condom use
118 10-1,000
Never
118 37
1-49%
70 22
50-99%
88 22
Always
62 20
Women with regular partner 192 60
Number of regular partners 1,7 1-8
Pregnant 16 5
No contraception use 171 54
Practice vaginal douching 223 82
After every intercourse
With
209 66
Water
27
9
Water and soap
158 50
Other (OMO, Dettol)
Use of lubricants
15
5
None
240 75
Vaseline
64 20
Saliva
9
3
Other
5
2
Sex during menses 71 23
Anal intercourse 45 14
Use alcohol
Medical (STD) history
158 50
Vaginal discharge
70 22
Genital ulcer
11
4
Abdominal pain
92 29
Genital itch
39 12
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 84

Water alone or water and soap were the fluids most commonly used for vaginal douching.
The use of lubricants, in most cases Vaseline, was practiced by 24% of the women. High-risk
sexual behavior was common. Anal intercourse was practiced by 14% of women and sex
during menses by 23% women. Almost half of the women reported a past history of STIS.
Alcohol use was reported by 50% and intravenous drug use by 9 (3%).
Overall, 80 (27%) women were HIV-1 infected. The prevalence of bacterial vaginosis was
46%, trichomoniasis 13%, gonorrhea 8%, chlamydia 7%, syphilis 6% and clinical ulcer 2%.
Squamous intraepithelial lesion (SIL) was detected in 10 cases (3%) of whom 5 were lowgrade
SIL (LGSIL), and 5 were high-grade SIL (HGSIL). The STD prevalence at the time of
screening for HIV-positive and HIV-negative women is shown in Table 2. Only T. vaginalis
was significantly more prevalent in HIV-positive women. HIV-positive women had more
bacterial vaginosis and more gonorrhea but these differences did not reach statistical
significance. There was a trend to increased CIN prevalence in HIV-1 infected women.
Table 2 : Sexually transmitted infection prevalence among HIV- positive and HIV- negative sex workers in Nairobi.
HIV- positive HIV- negative
n % n * % OR (95% CI) P
Bacterial vaginosis 83/186 45 38/70 54 1.5 (0.8-2.6) 0.2
Candida Gram stain 13/187 7 2/70 3 0.4 (0.1-1.8) 0.4
Trichomoniasis 20/219 9 18/80 23 2.9 (1.4-5.8) 0.005
Chlamydia
Gonorrhea
14/216 7 6/80 8 1.2 (0.4-3.2) 0.8
Culture
14/219 6 9/80 11 1.9 (0.8-4.5) 0.2
PCR
12/215 6 9/80 11 2.1 (0.9-5.3) 0.1
Syphilis 12/217 6 7/80 9 1.6 (0.6-4.3) 0.3
Clinical ulcer 5/219 2 1/80 1 0.7 (0.1-6.2) 0.6
CIN 4/219 2 5/80 6 3.6 (0.9-13.7) 0.06
*Denominators differ as not all tests were done (samples not taken, missing samples, missing results).
Age, age at first sex, years in Nairobi, number of partners per day, history of past STIs,
alcohol use, intravenous drug use, hormonal contraceptives, anal sex and sex during
menses were not significantly associated with HIV infection (Table 3). There was no
difference in mean duration of prostitution between HIV-positive and HIV-negative women.
The prevalence of HIV-1 infection was highest among women who had been involved in
prostitution for less than 2 years (38% vs. 24%, OR 1.6, 95% Cl 1.1-2.4, P=0.03). HIVpositive
women tend to charge less on average per sex act, not to have a regular partner, not
to practice vaginal douching but this was not statistically significant. HIV-positive women had
a significantly higher number of partners per week than HIV-negative women.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 85

Widows, divorcees and separated women were more HIV infected. Women who reported
always using condoms were 50% less likely to be HIV infected than women who reported no
condom use.
Table 3 : Characteristics among HIV- positive and HIV- negative sew workers in Nairobi.
HIV negative
% or mean
HIV positive
% or mean
OR (95% CI) P
Age 32.1 31.8 0.9 (0.9-1.4) 0.8
Years lived in Nairobi
Marital status :
11.4 10.3 1.0 (0.9-2.3) 0.3
Never married
43
33 0.8 (0.7-1.0) 0.06
Widowed/divorced/separated 57
68
Age at first sex 16.4 15.8 1.6 (0.9-1.2) 0.1
Duration prostitution 6.9 6.1 1.1 (0.6-1.4) 0.2
Partners/day 3.7 3.6 0.9 0.8
Partners/week 13.6 23.9 1.9 (0.6-20.9) 0.06
Charge per act
Condom use :
125 107 1.2 0.2
Never
36
38 1.7 (0.9-3.1)* 0.05*
1-49%
21
29
50-99%
21
23
Always
23
11
Regular partner 62 56 0.9 (0.7-1.2) 0.2
Use of oral contraceptives 11 10 1.1 (0.5-2.3) 0.5
Douching 71 66 0.9 (0.6-1.2) 0.5
Any lubricants used 28 18 1.6 (0.9-2.7) 0.04
Sex during menses 23 24 1.0 (0.9-1.2) 0.4
Anal intercourse 15 14 1.0 (0.9-1.1) 0.5
Any alcohol use 47 56 0.8 (0.7-1.1) 0.1
History of STD 41 44 0.9 (0.7-1.2) 0.4
*P value for categories «always» compared with «never». OR= odds ratios; CI= confidence intervals.
We compared the characteristics of the women according to the place of work (Table 4).
Women working in bars compared with women working at home were younger, had more
partners per week and used alcohol more often. They had a higher HIV prevalence but this
was not significantly different. Only 10 women were working in a nightclub and they were
more often single, charged more per sex act, had anal sex and sex during menstruation
more often, had a higher use of lubricants and used alcohol more often. Only one of them
was HIV infected.
To examine further the relation among risk factors for HIV infection, we performed logistic
regression analysis. All variables significantly associated with HIV in univariate analysis were
included in the model. These were marital status, partners per week, condom use, use of
lubricants, place of work, trichomoniasis and SIL.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 86

After logistic regression, only trichomoniasis (P=0.02) remains significantly associated with
HIV infection. The trend for association between SIL and HIV remains in the regression
analysis (P=0.07).
Discussion
Table 4 : Characteristics of the 272 women working at home or in bars according to place of work.
Home n=216
% or mean
Bar n=56
% or mean
OR (95% CI) P
Age 33 29 0.001
Single 40 50 1.4 (0.9-1.0) 0.1
Duration prostitution 6.8 6.0 0.3
Partners/day 3.7 3.7 0.9
Partners/week 14 27 0.04
Charge per act
Condom use :
110 120 0.6
Never
40
27 1.2 (0.8-1.7)*
0.5*
1-49%
19
36
50-75%
22
20
Always
19
18
Douching 69 77 1.5 (0.7-2.9) 0.2
Lubricants 22 27 1.3 (0.7-2.6) 0.3
Sex during menses 20 31 1.6 (1.0-2.6) 0.06
Anal intercourse 15 13 0.8 (0.4-1.7) 0.4
Alcohol use 43 66 2.2 (1.3-3.6) 0.001
HIV infection 25 29 1.0 (0.9-1.2) 0.3
*P value for categories «always» and «never». OR= odds ratios; CI= confidence intervals.
This paper describes the baseline data of the 318 sex workers screened during the first 7
months of the enrolment phase of a randomized placebo controlled trial with azithromycin to
reduce STI incidence. This cohort of female sex workers was set up in a slum area in
Nairobi.
The women in our study were initially mainly recruited through an existing network of peer
leaders. As it became clear that some FSWs were not reached by these peer-leaders, other
ways of recruiting them were identified. As a result, women working in bars and clubs were
also enrolled later on in the study although at a slower pace. The women we enrolled form a
non-coherent group of sex workers with striking differences in relation to demography as well
as to risk behavior according to the place where they were working. Women carrying out
their activities at home were older, more often divorced or widowed and had been in
prostitution for a longer time.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 87

They did not report much condom use but on the other hand did not report other high-risk
sexual behavior very often. A smaller group of women worked out of bars. These bars are
located within the slum area and are often not more than a shabby construction where locally
brewed beer is served. The women working here were younger, had more sex partners and
reported a little more high-risk sexual behavior. This group of women shows the highest rate
of HIV infection. We recruited only 10 women working in clubs. Although they have different
characteristics, it is difficult to draw conclusions given the low numbers. The assumption can
be made that the clients of these 3 subgroups of FSWs are quite different, which may also
account for different rates in condom use as well as different HIV prevalence rates.
However, as no questions on client characteristics were included in the questionnaire, this
cannot be confirmed. It is clear however that adapted target interventions will be needed to
reach these different women.
Most of the FSWs in our study, especially the ones working at home, have been reached by
a community-based program through peer leaders that has been in place for several years.
This program focused on health education towards behavioral change as well as on
alternative income generation. This might explain the differences observed in this cohort as
compared with the sex workers of earlier established cohorts in Nairobi, before health
education programs had been initiated. The most obvious difference is the rather low STI
prevalence rates in our cohort. The earliest cohort of FSWs in the Pumwani slum area in
Nairobi showed an HIV prevalence rate of 67% (Plummer 1991). We found an overall
prevalence rate of 27%. Although this rate is higher than what was observed among
pregnant women in Nairobi (Jackson 1999), it corresponds to the prevalence rate among
women with vaginal discharge in the same area (Fonck 2000). The pattern of HIV infection in
relation to duration of prostitution, with decreasing HIV prevalence as duration of sex work
increases, as we observed, has been described before (Fowke 1996, Simonsen 1990).
Selection bias is a possible explanation of this finding. The women who became rapidly
infected have progressed towards immune deficiency and possibly death. The other subset
of women might be resistant to infection. This was first described by Fowke et al. in another
prostitute cohort in Nairobi.
Also the prevalence rates of the other STD were surprisingly low in our cohort. In a similar
cohort of sex workers in the Nairobi area in 1985, gonorrhea was found in 50%, chlamydia in
25%, genital ulcers in 28% and syphilis in 32% (Simonsen 1990). We found respectively 8%,
7%, 2% and 6%. No H. ducreyi was detected in the actual cohort while 13% of the sex
workers had culture proven chancroid in 1985. Studies among sex workers in other parts of
Africa have also reported much higher STI prevalence than what we found (Ramjee 1998,
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 88

Diallo 1998, Ndoye 1998, Germain 1997). This finding may reflect the general observed
decline of STIs in Nairobi. Temmerman et al. described a significant reduction of syphilis
prevalence since 1995 among pregnant women in Nairobi. Similarly, Moses et al. reported a
decline in STD syndromes among clinic attendees in primary health care centers in Nairobi in
1998. The authors believe that this decline is a real one and suggest it to be the result of
prevention and intervention programs in Nairobi.
The women in our cohort reported some condom use in 63% (while 37% reported no
condom use at all). In the earliest cohort in Nairobi, only 8% of the women reported some
condom use (Ngugi 1988). However, after the introduction of health education and condom
promotion in this cohort, the use of condoms increased dramatically resulting in over 60% of
the women reporting some condom use. This most obvious explanation of this high condom
use might be the existence of the community-based health education program. However, this
cannot explain why the women working at home, or the group best reached by the program,
reported less condom use than the other women. Mass education campaigns of the National
AIDS/STD Program might possibly also have had a positive influence on behavior change
within this community. Better economic factors and hence the ability to purchase the
condoms might also play a role. Another factor might be the difference in characteristics of
the clients, with the men frequenting the lower socio-economic sex workers refusing
condoms more often. A combination of several factors is the most plausible explanation for
this occurrence.
The women in our study reported sex during menses and anal intercourse in 23% and 14%
respectively. This is in contrast with results from the earliest cohorts of FSWs in Nairobi,
where these risk behaviors were not reported at all (Plummer 1991). In the later established
cohort only small numbers of women admitted to practicing anal and oral sex as well as sex
during menses (Ngugi 1996). In our study, these sexual behaviors were not associated with
more HIV infection probably because the women who practice them use condoms more
often.
We found a strong correlation between T. vaginalis infection and HIV infection. Several
authors have found an association between non-ulcerative STD and HIV (Wasserheit 1992,
Laga 1993). The action by which these STDs increase the susceptibility to HIV have not
been defined, but potential mechanisms include: increased number of HIV target cells in the
genital tract (Levine 1998); disruption of the normal epithelial barrier (Kiviat 1990); and
reduction of protective T-helper and cytotoxic T-lymphocyte function (Mazzoli 1997).
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 89

The low prevalence of the other STDs in our study might explain the failure to demonstrate
associations between those STDs and HIV.
The only other correlation with HIV we found was with CIN. This did not reach statistical
significance probably due to the small numbers. It has been demonstrated that women with
HIV are at greater risk for the development of lower genital tract neoplasia than are HIVnegative
women. Among HIV-positive women, those who are more severely
immunosuppressed appear to be at higher risk (Abercombie 1998). We did not have
information on the status of immonosuppression among the women in our study, and hence
cannot confirm this relation. However, as these women were still active as sex workers, it
can be assumed that their health status was relatively good, which might further explain the
failure to demonstrate a significant association.
The majority of the women practiced vaginal douching and most of them used water only.
Findings from other studies indicate a relation between douching with commercial antiseptics
and lower HIV prevalence although they found that douching with non-commercial
preparations was associated with a higher HIV prevalence (Gresenguet 1997). We equally
found that vaginal douching was associated with bacterial vaginosis. In a cross-sectional
study in Uganda, bacterial vaginosis was associated with increased HIV infection among
younger women (Sewankambo 1997). In a subsequent longitudinal study, bacterial vaginosis
was significantly associated with antenatal and postnatal HIV seroconversion (Taha 1998).
Hence, we may assume that the lower HIV prevalence among women who douched in our
study is probably not a true one and should be confirmed once higher numbers of women are
enrolled in the study.
In summary, we have identified a large population of HIV-uninfected sex workers who report
high-risk behavior. It appears feasible to access these women with prevention programs,
including a clinical trial. We also found that sex workers are not a coherent group and that
different prevention strategies for different subgroups are needed in order to be effective.
Health education programs, either at the micro or national level, seem to have had a positive
impact on behavior change among FSWs, resulting in lower STI prevalence rates.
Therefore, in order to reach the first objective of our study, we will have to consider
increasing either the sample size or the duration of follow-up. The negative finding of our
study however is the persistent low condom use despite intensive health education. This
underlines the need for continuous adapted health education programs among high-risk
groups as well as among the general population, especially among men.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 90

6.4. Sexually Transmitted Infections and Vaginal Douching in a
Population of Female Sex Workers in Nairobi, Kenya.
Summary
Published in Sexually Transmitted Infections 2001; 77:271-275
by Fonck K, Kaul R, Keli F, Bwayo J, Ngugi E, Moses S and
Temmerman M.
This study was undertaken to assess the association between vaginal douching and sexually
transmitted infections (STIs) among a group of female sex workers (FSWs) in Nairobi,
Kenya. This study was part of a randomized, placebo controlled trial of monthly prophylaxis
with 1g of azithromycin to prevent STIs and HIV infection in a cohort of Nairobi FSWs.
Consenting women were administered a questionnaire and screened for STIs. The
seroprevalence of HIV-1 among 543 FSWs screened was 30%. HIV infection was
significantly associated with bacterial vaginosis (BV), trichomoniasis, gonorrhea, and the
presence of a genital ulcer. Regular douching was reported by 72% of the women, of whom
the majority inserted fluids in the vagina, generally after each sexual intercourse. Water with
soap was the fluid most often used (81%), followed by salty water (18%), water alone (9%),
and a commercial antiseptic (5 %). Douching in general and douching with soap and water
were significantly associated with bacterial vaginosis (P = 0.05 and P = 0.04 respectively).
There was a significant trend for increased frequency of douching and higher prevalence of
BV. There was no direct relation observed between douching and risk for HIV infection or
other STIS. The widespread habit of douching among African female sex workers was
confirmed. The association between vaginal douching and BV is of concern, given the
increased risk of HIV infection with BV, which has now been shown in several studies. It is
unclear why we could not demonstrate a direct association between douching and HIV
infection. Further research is required to better understand the complex relation between
douching, risk for bacterial vaginosis, and risk for HIV and other STIs.
Introduction
Heterosexual intercourse is the major route of transmission of HIV in sub-Saharan Africa.
The role of ulcerative and non-ulcerative sexually transmitted infections (STIS) in facilitating
the transmission of HIV is well established. However, the role of local genital tract factors in
HIV transmission is less clear.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 91

There is some evidence that frequent vaginal douching may increase women's susceptibility
to sexually transmitted agents, through modification of the vaginal flora. Several studies have
suggested that vaginal douching may increase risk for cervical infections (Scholes 1998,
Critchlow 1995, Joesoef 1996, Van De Wijgert 2000). Others have shown that douching
predisposes women to pelvic inflammatory disease (PID), with a twofold increased risk of
PID associated with douching (Scholes 1993, Foxman 1998, Wölner-Hansen 1990, Aral
1992). Genital tract infections such as cervicitis and PID have, in turn, been identified as a
risk factor for HIV acquisition. Hence, vaginal douching may indirectly have a facilitating role
in the heterosexual transmission of HIV. In addition, vaginal douching has been reported to
be associated with reduced fertility (Baird 1996). Finally, douching has been suggested as a
risk factor for cervical cancer, although the evidence for this has been inconsistent (Zhang
1997).
The insertion of various substances into the vagina is a common global practice. Intravaginal
substances, largely through their astringent properties, are perceived to enhance sexual
pleasure in many areas of sub-Saharan Africa. In Zimbabwe, 87% of clinic attendees and
nurses interviewed reported this habit (Runganga 1992). The use of vaginal agents for the
treatment of vaginal discharge was reported by 30% of women in a study in the Central
African Republic (Gresenguet 1997) and is also common in Malawi (Dallabeta 1995).
This study was undertaken to assess the association between vaginal douching and sexually
transmitted infections among a group of female sex workers (FSWs) in Nairobi, Kenya, as
part of a large randomized trial of monthly azithromycin prophylaxis to prevent STIs and HIV
infection in a cohort of Kenyan sex workers. The study design and baseline and preliminary
findings have been published elsewhere (Fonck 2000, Moses 2000).
Methods
As part of a randomized, placebo controlled trial of monthly prophylaxis with 1 g of
azithromycin to reduce the incidence of STIs and HIV infection among female sex workers in
a Nairobi slum area, a structured questionnaire was administered to all women presenting for
screening into the study. Written, informed consent was obtained from all women. The
questionnaire gathered data regarding vaginal douching, which was defined as the insertion
of any liquid into the vagina, and also included data on demographics, current sexual
behavior (client numbers, condom use, and sex practices), and reproductive and medical
histories over the past year.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 92

A full physical examination, including speculum examination and laboratory STI testing, was
performed at the screening visit. Data on the first 543 women screened are presented here.
The study was approved by ethics review committees of the University of Nairobi and the
University of Manitoba. A detailed description of the specimens taken and the laboratory
tests performed has been published elsewhere (Fonck 2000).
Data were entered into a database in Microsoft ACCESS (Office 97 format) and analysis
performed after export into SPSS for Windows, version 8.0 (SPSS Inc, Chicago, IL, USA). In
univariate analysis, the odds ratios and 95% confidence intervals were used for the
measurement of associations between proportions. Comparisons were made using
Pearson's χ 2 , Fisher's exact test, and χ 2 test for trend. Student's t test was used for
comparison of means. Logistic regression models were developed to analyze risk factors for
HIV and BV. These models used input variables that were associated with HIV or BV in
univariate analysis in this cohort, or which have been reported in other cohorts.
Results
Between May 1998 and August 1999, 543 female sex workers were screened. Three women
declined examination after interview and were excluded. The mean age was 30 years (SD
7.8 years). Overall, 161 women (30%) were HIV-1 seropositive. The prevalence of bacterial
vaginosis was 49%, candidiasis 10%, trichomoniasis 16%, gonorrhea 10%, chlamydia 9%,
syphilis 6%, clinical ulcer 1%, and CIN 2%. Only one woman presented with genital warts.
Demographic characteristics, sexual behavior, and medical history of HIV positive and HIV
negative women are compared in Table 1.
There was no association between age, marital status, duration of prostitution, number of
partners, and HIV serostatus. HIV positive women had had their first sexual experience at
younger age than HIV negative women, charged less per sexual act, used condoms less
often, were less likely to report a regular partner, and used alcohol more frequently. Use of
oral contraceptives, douching, sex during menses, anal intercourse, and history of STIs were
not associated with HIV serostatus. On multivariate analysis only age at first sex and alcohol
use remained significantly associated with HIV seropositivity (P=0.02 and P=0.05,
respectively).
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 93

Table 1. Characteristics of HIV- positive and HIV- negative sex workers.
HIV- negative
(n = 373)
HIV- positive
(n = 161)
OR (95% CI) P
Mean age (years) 30.4 30.9 0.4
Mean number of years lived in Nairobi
Marital status
10.6 9.9 0.4
Never married
Widowed/divorced/separated
50 %
50 %
46 %
54 %
0.9 (0.7-1.1) 0.2
Mean age at first sex (years) 16.3 15.6 0.005
Mean duration of prostitution (years) 6.1 6.2 0.8
Mean number of sex partners/day 3.9 4.0 0.5
Mean number of sex partners/week 13.9 19.1 0.08
Mean charge per act ( Kenya shillings)
Condom use :
118 93 0.02
Never
34 % 35 % 1.9 (1.0-3.6) 0.04
Sometimes
47%
55% 2.1 (1.2 –3.9) 0.01
Always
19 % 10 % 1.0 (reference) N/A
Regular partner 57 % 47 % 0. 8 (0.6-1.0) 0.02
Use of oral contraceptives 31 % 31 % 1.0 (0.8-1.3) 0.5
Sex during menses 23 % 26 % 1.1 (0.8-1.5) 0.2
Anal intercourse 15 % 16 % 1.1 (0.8-1.5) 0.4
Alcohol use 55 % 66 % 0.7 (0.5-0.9) 0.01
History of STIs 28 % 27 % 1.0 (0.8-1.4) 0.5
History of douching 72 % 73 % 1.0 (0.7-1.6) 0.9
STI prevalence at the time of
screening stratified by HIV
serostatus is shown in Table 2.
Bacterial vaginosis,
trichomoniasis, gonorrhea, and
the presence of a genital ulcer
were significantly more common
among HIV positive women.
CIN was three times more
prevalent in the HIV seropositive
group.
Table 2. STI prevalence among female sex workers by HIV serostatus*.
HIV- negative HIV- positive
n % n % OR (95% CI) P
Bacterial vaginosis 138/303 46 73/129 57 1.5 (1.0-2.3) 0.05
Candida 28/303 9 11/130 9 0.9 (0.4-1.9) 0.9
Trichomoniasis 43/369 12 41/156 26 2.7 (1.7-4.4) < 0.001
Chlamydia 30/327 9 11/124 9 1.0 (0.5-2.0) 1.0
Gonorrhea 30/370 8 22/156 14 1.9 (1.0-3.3) 0.04
Syphilis 23/369 6 8/159 5 0.8 (0.3-1.8) 0.7
Clinical ulcer 2/373 1 5/161 3 6.0 (1.1-31.0) 0.03
CIN 4/373 1 5/161 3 3.0 (0.8-11.2) 0.2
*
Denominators differ because of missing specimens.
Of the 543 women interviewed, 392 (72%) gave a history of douching. Most of the women
who practiced douching reported doing so after each sexual intercourse (91%) and most of
the women douched with water and soap (81%). Water mixed with salt was used by 18%,
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 94

water alone by 9%, a commercial antiseptic by 5%, and washing powder was used by 1%.
The majority of women who reported douching did so more than once per day (93%).
There was no difference in age, marital status, age at first sexual intercourse, place of work
(home, bar, or club), and duration of prostitution between women who douched and those
who did not (data not shown). There was a stepwise association between condom use and
douching: women who sometimes or always used condoms were 1.4 and 2.5 times more
likely to douche, respectively, than women who never used condoms (P=0.003). Although
women who douched had significantly more partners per day (4.1 vs. 3.6; P=0.01) and
practiced anal sex more often (19% vs. 5%; P

A significantly higher prevalence of bacterial vaginosis was found among women who
douched (OR 1.5, 95% CI 1.0-2.3, P=0.05). This association was also found in the subgroup
of women douching with soap (OR 1.6, 95% CI 1.0-2.5, P=0.03). The correlates of bacterial
vaginosis are shown in Table 4. Bacterial vaginosis was more prevalent among women who
used alcohol more often, among women with trichomoniasis, chlamydial infection, and HIV
infection. Bacterial vaginosis was less prevalent among pregnant women and women with a
history of past vaginal discharge. When the variables associated with bacterial vaginosis in
univariate analysis (P

in the group douching once per day, and 14% among those douching less than once daily
(P=0.05). More frequent douching was also associated with higher pH (P=0.05). A lack of
lactobacilli was detected significantly more often in women douching with salt (73% versus
47%, OR 0.3, 95% CI 1.1-5.0, P=0.04). There was no difference in lactobacilli detection
observed among women douching with other products.
Discussion
The current study confirms that vaginal douching is a widespread practice among African
female sex workers, as previously described. In our study, 72% of the women reported
vaginal douching. Vaginal douching is also widely practiced among pregnant women and
women attending STD clinics in Africa (Dallabeta 1995, La Ruche 1999). The use of vaginal
products for the treatment of vaginal symptoms as well as the use of vaginal agents to
achieve a tightening effect is widespread (Dallabeta 1995, Mann 1988, Williams 1993). The
specific preparations used for douching vary according to local cultural factors. In several
African settings, herbs and dry leaves are used for treatment of vaginal infections or for a
‘tightening effect’ to achieve dry sex (Gresenguet 1997, Brown 1993). Although the question
was not specifically asked in our study, it appears that personal hygiene was the main
reason for douching, as most women did so after each intercourse. Most women used water
alone or a combination of water and soap. Commercial products were rarely used. The
women in our study belonged to a low socio-economic group, and hence may not have been
able to afford rather expensive commercial products.
The STD and HIV rates in this population are similar to other African cohorts, except for the
CIN rate of 2%, which is lower than reported from most other studies, even in low risk groups
in east Africa (Temmerman 1999). Health-seeking behavior and previous screening and
treatment of cervical lesions might explain these low rates of CIN.
We describe a significant association between douching (with soap and water) and an
increased prevalence of bacterial vaginosis. More frequent douching was also significantly
associated with more prevalent bacterial vaginosis. This is consistent with results from other
studies. A case-control study has shown that genital hygiene accounts for a twofold increase
in the risk of bacterial vaginosis (Rajamanoharan 1999). Although the cross sectional nature
of these studies precludes the establishment of a cause and effect relation (for instance,
women with BV symptoms may douche more often), the finding in a prospective cohort study
that acquisition of bacterial vaginosis is associated with douching (Hawes 1996) makes a
causal relation more likely.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 97

There is now considerable evidence that the presence of bacterial vaginosis has a role in the
acquisition of HIV (Schmid 2000). Hence, vaginal douching may indirectly facilitate the
heterosexual transmission of HIV infection. We did not find a direct association between
douching and HIV prevalence, but other factors associated with douching in our cohort may
have acted to reduce HIV risk. Women who reported douching were more likely to report
using condoms all the time. Vaginal douching appeared to be used as a form of personal
hygiene, and it seems reasonable that women more concerned with hygiene might also be
more likely to use condoms. The association between douching and BV would still be
apparent, but this increased condom use would bias our ability to detect any effect of BV on
enhanced HIV acquisition. As has been seen in previous studies on this subject, there
appears to be a complex relation between HIV infection and vaginal douching. In a study in
the Central African Republic, Gresenguet et al. found a positive association between HIV and
the use of vaginal agents for the self-treatment of discharge and itching. In women without
vaginal symptoms, however, the use of vaginal agents was not associated with HIV infection.
Dallabeta et al. found an increased prevalence of HIV infection among women using noncommercial
medicines in Malawi. The potential mechanisms whereby vaginal douching could
possibly enhance HIV transmission are twofold: (1) through irritation of the vaginal mucosa,
thus promoting the proliferation of lymphocytes which are target cells for HIV (Peterman
1990); and (2) through dehydration of the vaginal mucosa, rendering the vaginal epithelium
more vulnerable to local trauma (Irwin 1993).
In summary, frequent douching seems to be a common practice among female sex workers
in Africa, and we found a significant association between douching and bacterial vaginosis,
but not with other STIs or HIV infection. The association between vaginal douching and
bacterial vaginosis is of concern, given the increased risk of HIV infection with BV, which has
now been shown in several studies. It is unclear why we could not demonstrate a direct
association between douching and HIV infection, but this may be due to a complex
interaction with behavioral and other factors associated with douching. Further research is
required to better understand the complex relation between douching, risk for bacterial
vaginosis, and risk for HIV and other STIs.
PREVENTION OF SEXUALLY TRANSMITTED INFECTIONS INCLUDING HIV 98

7.1. Summary of key findings
C HAPTER 7
CONCLUSIONS
In Chapter 3 we described that the major complaint among women seeking medical care at
the STI referral clinic in Nairobi was vaginal discharge, and candidiasis the most commonly
found pathogen. HIV prevalence rate was 29% in this group and HIV infection was strongly
associated with sexual behavior, especially number of sexual partners. Most of the women
however reported low-risk sexual behavior. Condom use was rare. On the other hand,
analysis of routinely collected data on syphilis serology indicates a significant decline in
syphilis among pregnant women in Nairobi between 1993 and 1997.
The healthcare-seeking behavior studies (Chapter 4) showed important gender differences
in STI health-seeking behavior. There are long delays between appearance of symptoms
and seeking medical care, especially for women. Knowledge on health in general and STI
specifically is poorer for women. Many patients sought care elsewhere before attending the
public health sector and spent considerable amounts of money, especially men. Men more
so than women favor non-public health settings for STI treatment. Both studies showed that
men have more partners and have more extra-marital affairs than women. Women on the
other hand have more unprotected sex while symptomatic, mostly with their regular partner.
Condom use was generally low, even during symptomatic STI episodes.
In Chapter 5 we analyzed the validity of the vaginal discharge algorithm currently in use in
Kenya. The most commonly found problem among women with vaginal discharge was
candidiasis and trichomoniasis. However, also CT and NG were highly prevalent especially
among non-pregnant women, even at primary health care level. The actual algorithm for
vaginal discharge in use in Kenya did not perform very well. Introduction of the risk-score
would result in higher sensitivity, but at a higher overall cost.
In Chapter 6 we looked at different strategies for STI prevention. Routinely collected data on
syphilis screening in pregnancy revealed a RPR seroprevalence rate of 3.4%. Almost all
pregnant women registered were screened (96%), and 91% of all positive were treated.
Partner notification and treatment was 53%, which was twice as high as for non-pregnant STI
Chapter 7 99

patients. The analysis revealed high levels of false positive test results despite intensive
training and supervision. The cost per true case of congenital syphilis averted was estimated
at $80. At the main maternity hospital in Nairobi, syphilis seroprevalence rate among women
at delivery was 3%. Partner notification and treatment at delivery was only 36%. In a cohort
of female sex workers in a Nairobi slum area, we found relatively low levels of STI, including
HIV. Condom use was higher than expected. Vaginal douching in this cohort was highly
prevalent, and douching with soap and water was associated with BV, although we did not
find an association with HIV.
A number of issues arise from the results of our studies:
- How to explain the high STI prevalence rates in low-risk groups, while overall
syphilis rates decline and STI prevalence among FSW are relatively low;
- Can we apply the “Fransen-Piot” model and draw conclusions from it;
- What about screening and treatment versus mass treatment.
7.2. Core groups, bridging population and general population
We have demonstrated that STIs including HIV are highly prevalent in Nairobi, and this not
only in the population groups considered as being at high risk. The so-called population at
low risk (married and/or pregnant women), is showing high prevalence rates, indicating that
the epidemic has already spread from the core groups to the general population. Most of the
women enrolled in our studies can be considered at low-risk as they are in a stable
relationship and declare not having multiple sex partners. Obviously, the behavior of the
male partner (bridging group) plays a critical role in the spread of STI including HIV in this
population. We confirmed the behavior of those male partners who openly admit having
multiple sexual partners. Although it is known that women tend to underreport the number of
their sex partners while men tend to do the opposite, the number of sex partners reported by
men in our studies is quite high. A small proportion of men in our studies also admit paying
money for (commercial) sex hence having sexual contact with the core group. These findings
suggest that the core groups still play a role in STI transmission in Nairobi but that
interventions in this group alone will not be able to control the epidemic.
On the other hand, the STI prevalence rates that were observed in a newly established
cohort of commercial sex workers were relatively low. This may be a reflection of impact of
intervention activities, aiming at behavior change, in this group over the past years.
CONCLUSIONS 100

In contrast with high overall STI prevalence rates observed among men (self-reported) as
well as women in our various studies, we demonstrated a declining syphilis seroprevalence
rate among pregnant women over the past years in Nairobi. One explanation might be the
massive use of antibiotics among the general population in Nairobi, as by anecdotal
evidence. Antibiotics are freely available and sold by vendors, in markets and in pharmacies
without prescription. But this would not explain the high prevalence rates observed for the
other STIs. Another explanation of this observed downward trend in syphilis serology may be
the relative effective partner notification and treatment for syphilis during pregnancy, as we
described.
However, trends in reported STI cases should be interpreted with caution. Serwadda et al.
(1996) indicated that an observed decrease in HIV prevalence among pregnant women of
2% resulted from a decrease of 4.3% owing to mortality and net out-migration and an
increase of 2.3% owing to new infections. The declining syphilis seroprevalence rates we
observed might have been confounded by several factors: lack of information on patterns of
fertility, on age groups and on pattern of antenatal clinic use.
An important determinant of STI transmission dynamics is the duration of the infection in a
sexually active individual. In our studies, people with symptoms of STI waited long before
attending a health facility. It has been shown that overall women are more likely than men to
seek health care, however care for STIs appears to be an important exception for several
reasons. Women infected with an STI are far more likely to be asymptomatic. When
symptoms do occur, these are generally less clearly attributable to STIs. But even for
women who do recognize the symptoms, presentation to an STI clinic may be too stigmaladen.
An additional issue related to healthcare-seeking involves confidentiality. These
concerns may constitute a serious obstacle to seek medical care.
Condom use was found to be low in general but even more among women. Mathematical
modeling projections showed that the strategy of increasing condom use among women in
steady relationships is the least effective in reducing the incidence of HIV and this in different
profiles of sexual behavior and for different levels of condom use tested. These results are in
line with earlier ones based on more stylized models stressing the importance of reaching
high-risk groups, or the core, in STI prevention. Inconsistent condom use delays but does
not prevent HIV transmission in discordant couples (Van Vliet 2001).
Studies have suggested that one risk factor associated with recurrent STI among women is
continued sexual contact with a partner who failed to be treated (Fortenberry 1995). Another
CONCLUSIONS 101

contributory factor may be that male partners fail to inform other partners about the need for
treatment. We showed relatively high rates of partner notification among pregnant women
but this was considerably lower among non-pregnant women.
We confirmed that circumcision of the male partner was associated with lower HIV infection
rates in their female partners. A meta-analysis conducted by Weiss et al. (2000) provides
conclusive evidence that male circumcision is associated with a reduced risk of HIV infection
in sub-Saharan Africa. There is compelling biological and epidemiological evidence
supporting a protective effect of male circumcision on the acquisition of HIV infection and
ulcerative STDs in men in sub-Saharan Africa (Quigley 2001). However, there are still many
unknowns. These relate to the mechanisms and the role of the foreskin in the acquisition of
HIV, the existence of confounders in the attribution of causality, and the expected effect of
male circumcision on HIV in different populations (Van Dam 2000). There is little experience
concerning the practicability, feasibility, acceptability, and cost-effectiveness of male
circumcision as an HIV intervention. The effect of male circumcision on male and female risk
behavior and condom use is not known, but behavioral changes related to circumcision
status that result in reduced protection and increased risk-taking could well reduce the
beneficial effect of male circumcision. It might therefore be premature to recommend male
circumcision in currently non-circumcising communities.
While we cannot fully explain the situation observed, behavior change may have played a
role. The relatively low STI prevalence rates among FSWs in our studies may at least in part
be due to interventions among sex workers. The importance of changes in sexual behavior
has been demonstrated by data suggesting that in a number of developed and developing
countries public health campaigns targeting at reducing the spread of STIs have been
effective in reducing the prevalence and incidence of these infections (Gerbase 1998). The
high levels of HIV infection among adolescents in our studies highlight the importance of
interventions targeted at young people and their partners.
While behavior change interventions, condom promotion, promotion of improved healthseeking
behavior for the general population are needed, interventions targeted towards the
high groups should continue. Results from simulation models indicate the merit of continued
focusing of interventions on high-risk groups irrespective of the pattern of sexual behavior,
even in epidemics that have already spread to throughout the general population (Van Vliet
2001), as is the case in Kenya.
CONCLUSIONS 102

7.3. The operational model of effectiveness of STI services (Fransen-Piot model)
We did not have the ambition to collect information on all steps of the model. However, our
studies should be considered complimentary to similar earlier data on Nairobi (Rao 1998)
and may help in identifying the barriers. Based on a population of about 2 million, the annual
number of STI cases occurring in Nairobi can be roughly estimated at 225.000. The
proportion of infected individuals becoming symptomatic in our study was not known.
Korenromp et al (2002) have estimated those proportions at 45% for males with NG, 11% for
males with CT, 14% for females with NG and 6% for females with CT in rural Uganda.
However, these estimates were inconsistent with previous direct estimates based on a US
cohort study.
Comprehensive data on the number of patients with STI seeking medical care was not
available. Based on routine surveillance data of the NCC clinics providing STI care, including
the STC, we can estimate the number of annual STI cases seen at those facilities at about
70,000 (Temmerman 1999). Our data indicate that a large part of the STI care is not sought
through public sector clinics, however that proportion is unknown. Data on the proportion of
STI patients seeking medical care who were correctly treated were not available. We tried to
answer this question for women attending with vaginal discharge, and concluded that 42% of
women with CT and/or NG would be correctly diagnosed at first visit. Since all drugs required
for the STI protocols were available during the study period, we can assume that correct
treatment was always administered. A recent study on quality of STI care by different
categories of service providers in Nairobi indicates that correct diagnosis and treatment
varies between 14% and 48% (Voeten 2001). Public clinics equipped for STI care performed
best in all aspects, whereas treatment was poorest in pharmacies and private clinics.
Information on compliance and effectiveness of treatment was not available. As we indicated
that only a minority of cases is treated at public health facilities, correct overall treatment can
therefore be estimated to be very low. Partner notification varies from 12% for STI patients at
the STC clinic to 53% for partners of syphilis seropositive women during pregnancy.
There are obviously too many unknowns to calculate the proportion of patients with STI that
are cured by health services. In Mwanza, where improved STI services are offered in all
health centers, the estimated overall cure rate ranged between 23 and 41% (Buve 2001).
Based on our findings, we can assume that this proportion is probably lower in Nairobi.
We tried to answer the question on the quality of diagnosis of vaginal discharge among those
women attending the health care facilities, by evaluating the performance of the algorithm in
CONCLUSIONS 103

use in Kenya. In a supplement of Sexually Transmitted Infections, published in 1998, seven
papers evaluating the performance of the WHO vaginal discharge algorithm to gonococcal
and chlamydial infections were presented. Some consistent findings emerged from these
studies, and were also observed in our study (Dallabetta 1998). The symptom of vaginal
discharge is neither sensitive nor specific as a predictor of cervical infection. Symptoms of
vaginal discharge are more likely a manifestation of a vaginal infection, candidiasis,
trichomoniasis or BV. An approach that assigns treatment for cervical infection to women
with either positive risk assessment or signs of cervical or uterine inflammation is relatively
sensitive, but so non-specific that the PPV is probably too low to warrant treatment for
cervical infection in most settings. The diagnostic accuracy of all the currently available
simple screening / diagnostic tools for gonococcal and/or chlamydial infections is poor. And
finally, the clinical skills of the providers appear to significantly influence the performance of
decision models when physical assessment is included in the model. Furthermore, one
algorithm does not fit all settings. In family planning clinics, vaginal discharge algorithms with
high specificity for diagnosis of vaginal infection would be favored. On the other hand, in an
STD clinic or sex worker clinic, high sensitivity is favored because the prevalence of cervical
infection is elevated and the risk of future transmission is often high.
Based on our finding of high prevalence rates of NG/CT cervicitis in women presenting with
vaginal discharge, it would be appropriate to administer treatment for NG and CT at first visit
at least for non-pregnant women. This however implies an increased overall cost due to the
price of the antibiotics. There is a great need for simple, cheap, and reliable screening tests
for gonorrhea and chlamydial infection. This would allow screening at antenatal, family
planning, or maternal and child health clinics, and better management for symptomatic STI
patients in developing countries. Rapid tests for C. trachomatis and N. gonorrhoeae are on
the market however have not been evaluated. There is hope that rapid diagnostic tests of
proved value will be available within a few years (Mabey 2001).
We have been able to identify some obvious weaknesses at different levels that are useful to
indicate where it might be appropriate to focus interventions. Health-seeking behavior of
those individuals aware of their disease is poor, especially for women. Education and/or
counseling on appropriate healthcare-seeking implies awareness of potential signs and
symptoms of STIs and availability of adequate STI services, but both are generally found to
be weak (WHO 1999). Innovative strategies for improved health-seeking behavior are
therefore needed. To optimize STI health care, STI services must also be offered as part of
routine primary health care by both public and private sector providers.
CONCLUSIONS 104

7.4. Screening and/or mass-treatment
Since 1992 the prevalence of syphilis in pregnant women has been monitored in Nairobi. The
seroprevalence rate of syphilis among pregnant women in Nairobi has decreased
significantly in recent years. From a seroprevalence of 7.2% in 1994, over 7.3% in 1995, to
4.5% in 1996, 3.8% in 1997, we showed a syphilis serology among pregnant women of 3.0%
in 1998, and this both at the antenatal clinics and at the maternity hospital (Jenniskens
1995). Compared to the early years of the decentralized program, the quality of the syphilis
control program has improved considerably. For instance, partner referral and treatment at
antenatal clinics was 50% in 1992-93 and was as high as 75% in 1997-1998. However,
partner referral at postpartum reached only 36%, indicating that the well being of the unborn
child is an important determinant in partner notification. We showed in another study that
antenatal treatment of RPR positive women significantly improved pregnancy outcome but
the risk of adverse outcome remained 2.5-fold higher than the risk observed in uninfected
mothers (Temmerman 2000). Adverse pregnancy outcome was also related to lack of
partner treatment during pregnancy (Gichangi 2000). Presumptive STI treatment during
pregnancy resulted in reduced rates of STI as well as in improved neonatal outcome
(Gichangi 1997, Gray 2001).
However, we also showed that routine syphilis screening in this low seroprevalence rate
situation, is expensive and its effectiveness doubtful. Indeed, the proportion of patients with a
positive test result who actually have the disorder (PPV) is directly related to the prevalence
of the disease within the population for any given sensitivity and specificity of a diagnostic
test. This means that for a given test, the lower the prevalence of the disorder in a population
the greater the number of false positives. On the extreme, as the prevalence of a disease in
a population approaches zero, the PPV of any diagnostic test also approaches zero, a
concept depicted graphically by Ryan et al. (1998).
Under certain circumstances, mass treatment of a selected population group within a
community or region can be contemplated, for instance in situations where prevalence rates
of STIs are high and where laboratory facilities are insufficient or absent or when dealing with
highly mobile populations (WHO 1986). Prevalence rates can be reduced rapidly with mass
treatment, but in many if not all instances some infected individuals within the community and
region are not treated, and reintroduction of infection from outside the community or region
remains likely. Modeling indicates that STI prevalence would, after substantial reduction,
return to baseline levels over 5-10 years (Korenromp 2000). High rates of re-infection as can
be expected for instance in sex workers would necessitate periodic treatment supplemented
CONCLUSIONS 105

with intensive behavioral change interventions. Mass treatment may, in the short run, have
an impact on HIV incidence comparable to sustained syndromic management. Mass
treatment combined with sustained syndromic management may be particularly effective
(Korenromp 2000).
Presumptive treatment (selective mass treatment), where the PPV would equal the
prevalence of infection and sensitivity would be 100%, is a potential strategy to treat STIs in
high prevalence populations. It has to be decided at what level of infection presumptive
treatment becomes a cost-effective as well as ethically and programmatically acceptable
intervention. We think that in the Nairobi setting, mass treatment for syphilis could also be
considered for all pregnant women and their partners, although the implications of treatment
depend on whether immunity is present (Garnett 1997). A randomized control trial of mass
STI treatment among FSWs in Nairobi is underway. The results of this trial will be useful for
further decisions on interventions in this population.
7.5. Limitations and drawbacks
There are several limitations to our studies. First, all of our studies were cross-sectional,
even the results we presented as part of the randomized controlled trial. Hence, although
presence of associations has been raised, the causal relationships cannot be clearly
determined by our studies.
Second, systematic error may have been introduced to some extent. Selection bias may
have invalidated the conclusions we made. The sampling strategy in our studies was not
always random but sometimes purposeful or convenience. Furthermore, we did not always
routinely collect information on the individuals who refused to participate and hence bias
cannot be excluded. Anther limitation to our studies was observation bias, as different gold
standards were used across the studies. As those test have different sensitivities and
specificities, prevalence rates of STIs might have been over- as well as underestimated and
results may therefore not be comparable between different studies. Lack of adequate quality
control in some studies may have resulted in misclassification of results. Inter-observer bias
may also have occurred in the study where various clinicians were involved. Interviewer bias
in the multi-center studies may also have limited our results. And finally, recall bias and
underreporting of risk behavior may have affected the validity of the data on sexual behavior.
We have tried to control for confounding factors as much as possible. However, some
residual confounding may still have been present.
CONCLUSIONS 106

As all studies were clinic based, the findings can therefore not be extrapolated to the general
population. There is indeed evidence that both men and women in Nairobi use other health
care services (private, traditional, religious, etc) and that this choice is not random. The
different populations in our studies are therefore not representative for the general population
in Nairobi, let alone Kenya. In addition, the target groups we used in the various studies are
not comparable with each other for instance pregnant women attending PHC and FSWs.
Despite these limitations and drawbacks, we think that our study results are important in
having showed an overall picture on STIs in Nairobi, and in having identified some major
problems and possible intervention for improved STI care in Nairobi.
7.6. Priorities for future research
As we showed that STI patients in Nairobi prefer to attend the non-public health sector for
treatment, it would be useful to study the existence and quality of STI diagnosis and
treatment in that sector. It has to be noted that this study has been done partly as a result of
our findings.
Based on our results, research into appropriate behavioral change interventions, especially
targeted to men, seems a priority. Adolescents are an important group for behavioral change
interventions, and strategies to reach this target group should be established.
Strategies for improved health-seeking behavior relating to STI are also needed, especially
targeted to women. Research on ways to improve partner notification seems also needed.
A study on syphilis screening at prenatal clinic level including gestational age of screening
and treatment, RPR titers, parity and information on prevalence of abortion, should be able to
complement our findings and constitute a more founded basis to decide on the future of the
screening program. In the meantime, continuous monitoring of the syphilis-screening
program and of trends in seroprevalence rates is necessary. The impact of mass treatment
on syphilis requires further investigation.
In order to improve treatment of women with vaginal discharge, appropriate screening tests
for the detection of NG and CT are still urgently needed. The evaluation of the rapid tests
already on the market should take place as soon as possible.
CONCLUSIONS 107

C HAPTER 8
FINAL REMARKS
STIs illustrate the complex interactions of bio-ecology, medicine, culture and politics. The
AIDS epidemic shattered expectations of a relative stable world of infectious diseases. AIDS
like syphilis in the past engenders powerful social conflicts about the meaning, nature and
risks of sexuality. STIs, including HIV/AIDS, are probably more than any other public health
problem linked to structural problems of poverty, malnutrition and underdevelopment.
Many societies are reluctant to openly address issues involving sex and sexuality and to
recognize the realities of the sometimes-widespread existence of pre- and extramarital
sexual intercourse. Both for STI control as well as for prevention and control of HIV infection,
it is essential that health education and promotion efforts be intensified and sustained to
achieve an urgently needed change in risk-taking behaviors, to maintain safe behaviors, and
to develop an environment that enables people to adopt and sustain safe behavior (Van
Dam).
STI control is one of the key components to HIV prevention and control. Intervention
programs that include condom promotion and behavior change would benefit both STI and
HIV control programs. Furthermore, the efforts against STI must reflect the diversity of their
causes and their associations. It is unlikely that any single control measure will reduce HIV
transmission sufficiently and therefore different strategies will have to be implemented
simultaneously.
In resource poor settings, STI control programs should be integrated as much as possible
with HIV/AIDS control programs, since they will assure that duplication and wastage of
scarce resources is avoided. Areas for coordination or integration include care, health
education and counseling, promotion of safer sexual behavior, provision of condoms and
evaluation.
Different levels of interventions for STI prevention and control are possible: individual,
couple-based, core-group, community based, and population based. Although a multi-level
approach is ideal, limited resources may constrain choice. Other factors may also contribute
to selection of the appropriate strategy, for example the stage of the epidemic. In later stages
Chapter 8 108

of the epidemic when prevalence is more widespread, interventions directed towards core
groups may remain cost-effective but may no longer be sufficient to rapidly contain the
epidemic.
In ensuring universal access to appropriate STI care programs, it should be recognized that
patients seek care from a mixture of public and private sources. Planning of a balanced and
comprehensive program will need to consider strengthening any health care providers that
are able to provide a quality service. This means that in efforts directed at provider
behaviors, such as guideline development, training and evaluation must extend to providers
in the private sector as well.
Lack of political will and commitment, resistance from religious leaders, the culturally
accepted habits and old-fashioned ideas are all factors that may have enhanced this rapid
spread of HIV. There is a direct connection between the willingness of leaders to discuss
HIV/AIDS prevention and the success of such programs. In recent years, political leadership
in the fight against STI/HIV has improved considerably in certain countries hardest hit by the
HIV epidemic creating an enabling environment for action against the epidemic and its
negative effects on development. Other key elements needed for an effective answer to the
HIV/AIDS epidemic are the following: 1) a concerted action to integrate the HIV/AIDS
problematic in the main development instruments; 2) massive mobilization of additional
resources; 3) a multi-sectorial approach; 4) intensification of interventions based on the best
possible data.
Continued financial support for prevention and control programs is a concern. All countries
must continue to make efforts to mobilize domestic resources from all sources. The
international community is called upon to assist developing countries in their efforts. A strong
case can be made for providing free treatment for STIs on public health grounds, in the same
way it has been made for other communicable diseases such as tuberculosis, but this would
need to be supported by the international donor community in the poorest countries.
Introduction of cost-recovery mechanisms, as in place in Kenya, should not discriminate
against STI patients and should not ignore people’s needs. More adapted waiver systems as
to assure access for all, included the poorest, should be institutionalized. The coordination of
activities must be enhanced. Measures such as joint programs can improve coordination and
ensure a more efficient use of scarce resources.
Female controlled prevention methods such as female condoms and microbicides are of
particular relevance to the STI and HIV prevention agenda. Women should be given the tools
FINAL REMARKS 109

to protect themselves and be empowered to use them. However, although use of female
condoms may be an option for FSWs, its use in stable couples may be limited. Furthermore,
no effective microbicides are available as of yet, although research is promising. Since
empowerment of women in linked to socio-cultural changes and may not happen overnight, it
is of utmost importance to increase men’s awareness and responsibility towards safer sexual
behavior. Simultaneously, improved partner notification and treatment of the frequently
asymptomatic female partners of men seeking STI care may be one of the very few feasible
strategies for the management of women with STI. The alarming reports of high STI
prevalence rates among young people, especially women, call for urgent action and
implementation of appropriate programs targeted towards behavior change.
FINAL REMARKS 110

R EFERENCES
Abercombie PD, Kom AP. Lower genital tract neoplasia in women with HIV infection.
Oncology 1998; 12:1735-9.
Abyad A. Cost-effectiveness of antenatal screening for syphilis. Health Care Women Int
1995; 16:323-328.
IARC Scientific Publication no. 63: Africa; Lyons: IARC 1984; 451-463.
Ahmed SM, Adams AM, Chowdhury M, Bhuiya A. Gender, socio-economic development and
health-seeking behaviour in Bangladesh. Soc Sci Med 2000; 51:361-71.
AIDS epidemic update: December 1999. Geneva: UNAIDS and WHO.
Alary M, Baganizi E, Guedeme A, et al. Evaluation of clinical algorithms for the diagnosis of
gonococcal and chlamydial infections among men with urethral discharge or dysuria and
women with vaginal discharge in Benin. Sex Transm Inf 1998; 74(Suppl 1):S44-9.
Allen S, Tice J, Van de Perre P, et al. Effect of serotesting with counseling on condom use
and seroconversion among HIV discordant couples in Africa. BMJ 1992; 304:1605-9.
Anderson M. Epidemiology of HIV infection: Variable incubation plus infectious periods and
heterogeneity in sexual activity. J R Stat Soc 1988; 151:66-93.
Anderson R, May R. Epidemiological patterns of HIV transmission. Nature 1988; 333:514-19.
Anderson R, May R. Transmission dynamics of HIV infection. Nature 1987; 26:137-142.
Andrus JK, Fleming DW, Harger DR et al. Partner notification: can it control epidemic
syphilis? Ann Intern Med 1990; 112:539-43.
Aral SO, Mosher WD, Cates W Jr. Vaginal douching among women of reproductive age in
the United States: 1998. Am J Public Health 1992; 82:210-4.
Asuzu MC, Rotowa NA, Ajayi IO. The use of mail reminders in STD contact tracing in Ibadan,
Nigeria. East African Med J 1990; 67:75-8.
Asuzu MC, Ogubanjo BO, Ajayi IO, Oyediran ABO, Osoba AO. Contact tracing in the control
of STD in Ibadan, Nigeria. Br J Veneral Dis 1984; 60:114-6.
Azeze B et al. Seroprevalence of syphilis amongst pregnant women attending antenatal
clinics in a rural hospital in northwest Ethiopia. Genitourin Med 1995; 71:347-50.
Baird DD, Weinberg CR, Voigt LF, Daling JR. Vaginal douching and reduced fertility. Am J
Public Health 1996; 86:844-50.
Bam RH, Cronje HS, Muir A, Gziessel DJ, Hoek BB. Syphilis in pregnant patients and their
offspring. Int J Gynecol Obstet 1994; 44:113-18.
Behets FM, Ward E, Fox L et al. Sexually transmitted diseases are common in women
attending Jamaican family planning clinics and appropriate detection tools are lacking. Sex
Transm Infect 1998; 74(suppl1):S123-7.
References 111

Bell G, Ward H, Day S et al. Partner notification for gonorrhoea: a comparative study with a
provincial and a metropolitan UK clinic. Sex Transm Inf 1998; 74:409-14.
Berkowitz G, Papiernik E, Epidemiology of preterm birth. Epidemiol Rev. 1993; 15:414-443.
Blanckhart D, Muller O, Gresenguet G, Weis P. Sexually transmitted infections in young
pregnant women in Bangui, Central African Republic. Int J STD AIDS 1999; 10:609-14.
Braddick M, Ndinya-Achola J, Mirza N, et al. Towards developing a diagnostic algorithm for
Chlamydia trachomatis and Neisseria gonorrhoeae cervicitis in pregnancy. Genitourin Med
1990; 66:62-5.
Brandt AM. The syphilis epidemic and its relation to AIDS. Science 1988; 239:375-80.
Brown RC, Brown JE, Ayowa OB. The use and physical effects of intravaginal substances in
Zairian women. Sex Transm Dis 1993; 20:96-9.
Brown WJ, Donohue JE, Axnick NW, Blount JH, Ewen NH, Jones OG. Syphilis and other
venereal diseases. Cambridge, Massachusetts: Harvard University Press; 1970,1–81.
Brunham R, Plummer F. A general model of sexually transmitted disease epidemiology and
its implications for control. Med Clin North Am 1990; 74:1339-1352
Buve A, Weiss HA, Laga M et al. The epidemiology of trichomoniasis in women in four
African cities. AIDS 2001; 15(suppl 4):S89-96.
Buve A, Weiss HA, Laga M et al. The epidemiology of gonorrhoea, chlamydial infection and
syphilis four African cities. AIDS 2001; 15(suppl 4):S79-88.
Buve A, Changalucha J, Mayaud P et al. How many patients with sexually transmitted
infection are cured by health services? A study from Mwanza region, Tanzania. Trop Med Int
Health 2001; 6:971-79.
Cameron D, Simonsen J, D’Costa L et al. Female to male transmission of human
immunodeficiency virus type 1: risk factors for seroconversion in men. Lancet 1989; ii:403-
407.
Cates W, Farley T, Rowe P. Worldwide patterns of infertility: is Africa different? Lancet 1985;
2:596-8.
Clottey C & Dallabetta G. Sexually transmitted diseases and human immunodeficiency virus:
epidemiologic synergy? Inf Dis Clinics of North America 1993; 7:753-770.
Cohen C, Sinei S, Bukusi E, Bwayo J, Holmes K, Brunham. Human leukocyte antigen class
II DQ alleles associated with trachomatis tubal infertility. Obstet Gynecol 2000; 95:72-7.
Colvin M, Abdool Karim SS, Connolly C, Hoosen AA, Ntuli N. HIV infection and
asymptomatic sexually transmitted infections in a rural South African community. Int J STD
AIDS 1999; 9:548-550.
Costa LJ, Plummer FA, Bowmer I, et al. Prostitutes are a major reservoir of sexually
transmitted diseases in Nairobi, Kenya. Sex Transm Dis 1985; 12:64-7.
References 112

Costello Daly C, Wangel AM, Hoffman IF, et al. Validation of the WHO diagnostic algorithm
and development of an alternative scoring system for the management of women presenting
with vaginal discharge in Malawi. Sex Transm Inf 1998; 4(Suppl 1):S50-8.
Costello Daly C, Maggwa N, Mati J, et al. Risk factors for gonorrhoea, syphilis, and
trichomonas infections among women attending family planning clinics in Nairobi, Kenya.
Genitourin Med 1994; 70:155-6 1.
Critchlow CW, Wölner-Hanssen P, Eschenbach DA et al. Determinants of cervical ectopia
and of cervicitis: age, oral contraception, specific cervical infection, smoking and douching.
Am J Obstet Gynecol 1995; 173:534-43.
Dallabetta G, Gerbase A, Holmes K. Problems, solutions, and challenges in syndromic
management of sexually transmitted diseases. Sex Transm Inf 1998; 74(suppl):1-11.
Dallabeta GA, Miotti PG, Chiphangwi JD, Liomba G, Canner JK, Saah AJ. Traditional vaginal
agents: use and association with HIV infection in Malawian women. AIDS 1995; 9:293-7.
Day S, Ward H, Ghani AC et al. Sexual networks and the control of gonorrhoea. Int J STD
AIDS 1998; 9:666-71.
Desormeaux J, Behets F, Adrien M et al. Introduction of partner referral and treatment for
control of sexually transmitted diseases in a poor Haitian community. Int J STD AIDS 1996;
7:502-6.
Diallo MO, Ghys PD, Vuylsteke B, et aL Evaluation of Simple diagnostic algorithms for
Neisseria gonorrhoeae and Chlamydia trachomatis cervical infections in female sex workers
in Abidjan, Côte d' Ivoire. Sex Transm Inf 1998; 74(Suppl 1):Sl06-11.
Dietrich M, Hoosen A et al. Urogenital tract infections in pregnancy at King Edward VIII
Hospital, Durban, South Africa. Genitourin Med 1992; 68:39-41.
Donders G, Desmyter J, Hooft P, Dewet GH. Apparent failure of one injection of benzathine
penicillin G for syphilis during pregnancy in Human immunodeficiency virus seronegative
African women. Sex Trans Dis 1996; 24:94-101.
Draper D, Donohoe W, Mortimer L, Heine RP. Cysteine proteases of Trichomonas vaginal
degrade secretary leukocyte protease inhibitor. J Infect Dis 1998; 178:815-819.
Duncan M et al. Seroepidemiological and socio-economic studies of genital chlamydial
infection in Ethiopian women. Genitourin Med 1992; 68:221-7.
Easmon C, Hay P, Ison C. Bacterial vaginosis: a diagnostic approach. Genitourin Med 1992;
68:134-8.
Engels H, Nyongo A, Temmerman M, et al. Cervical cancer screening and detection of
genital HPV infection and chlamydial infection by PCR in different groups of Kenyan women.
An Soc Belg Med Trop 1992; 72:53-62.
Fast MV, Nsanze H, D'Costa LJ, et al. Treatment of chancroid by clavulanic acid with
amoxicillin in patients with P-lactamase-positive Haemophilus ducreyi infection. Lancet 1982;
2:509 -511.
References 113

Fleming D and Wasserheit J. From epidemiological synergy to public health policy and
practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV
infection. Sex Transm Inf 1999; 75:3-17.
Fonck K, Kaul R, Kimani J et al. A randomised, placebo-controlled trial of azithromycin
prophylaxis to prevent STI and HIV-1 in Kenyan sex workers: study design and baseline
findings. Int J STD AIDS 2000; 11:804-11.
Fonck K, Kidula N, Jaoko W, et al. Validity of the vaginal discharge algorithm among
pregnant and non-pregnant women in Nairobi, Kenya. Sex Transm Infect. 2000; 76:33-38.
Fonck K, Mwai C, Rakwar J et al. Health-seeking and sexual behaviour of patients with
sexually transmitted diseases in Nairobi, Kenya. Sex Transm Dis 2001; 28:367-71.
Forsythe S, Rau B, eds. AIDS in Kenya: Socio-Economic Impact and Policy Implications.
Arlington, Va: AIDSCAP/Family Health International; 1996.
Fortenberry J et al. Recurrent sexually transmitted infections (STI) among adolescent
females (Abstract 172). XI th meeting of the International Society for STD Research. New
Orleans, LA, August 1995.
Fowke K, Nagelkerke N, Kimani J, et al. Resistance to HIV-1 infection among persistently
seronegative prostitutes in Nairobi, Kenya. Lancet 1996; 348:1347-51.
Foxman B, Aral SO, Holmes KK. Interrelationships among douching practices, risky sexual
practices, and history of self-reported sexually transmitted diseases in an urban population.
Sex Transm Dis 1998, 25:90-9.
Garnett GP, Aral SO, Hoyle DV, Cates W, Anderson RM. The natural history of syphilis:
implications for the transmission dynamics and control of infection. Sex Transm Dis 1997;
24:185-200.
Geissler PW, Nokes K, Prince RJ, Odhiambo RA, Aagaard-Hansen J, Ouma JH. Children
and medicine: self-treatment of common illnesses among Luo schoolchildren in western
Kenya. Soc Sci Med 2000; 50:1771-83.
Gerbase AC, Rowley JT, Heymann DHL, Berkley SFB, Piot P. Global prevalence and
incidence estimates of selected curable STDs. Sex Transm Inf 1998; 74(suppl1):S15.
Germain M Alary M, Guèdèmè A, et al. Evaluation of a screening algorithm for the diagnosis
of genital infections with Neisseria gonorrhoeae and Chlamydia trachomatis among female
sex workers in Bénin. Sex Transm Dis 1997; 24: 109-15.
Gichangi P, Fonck K, Kigondu C, et al. Partner notification of pregnant women infected with
syphilis in Nairobi, Kenya. Int J STD AIDS. 2000; 11:257-261.
Gichangi P, Ndinya-Achola 0, Ombete J, Nagelkerke N, Temmerman M. Antimicrobial
prophylaxis in pregnancy: a randomized, placebo-controlled trial with cefetamet-pivoxil in
pregnant women with a poor obstetric history. Am J Obstet GynecoL 1997; 177:680-684.
Goldman N, Heuveline P. Health-seeking behaviour for child illness in Guatemala. Trop Med
Int Health 2000; 5:145-55.
References 114

Gray RH, Wabwire-Mangen F, Kigozi G et al. Randomized trial of presumptive sexually
transmitted diseases therapy during pregnancy in Rakai, Uganda. Am J Obstet Gynecol
2001; 185:1209-17.
Green M. Public reform and the privatisation of poverty: some institutional determinants of
health seeking behaviour in southern Tanzania. Cult Med Psychiatry 2000; 24:403-30.
Gresenguet G, Kreiss J, Chapko M, Hillier S, Weiss N. HIV infection and vaginal douching in
Central Africa. AIDS 1997; 11:101-6.
Grosskurth H, Mosha F, Todd J, et al. Impact of improved treatment of sexually transmitted
diseases on HIV infection in rural Tanzania: a randomised controlled trial. Lancet. 1995; 346:
530-536.
Guinness LF, Sibandze S, McGrath E, et al Influence of antenatal screening on perinatal
mortality caused by syphilis in Swaziland. Genitourin Med 1988; 64:294-7.
Gunn RA, Harper SL. Emphasizing infectious syphilis partner notification. Sex Transm Dis
1997; 25:218-9.
Handsfield HH, Sparling PF. Neisseria gonorrhoeae. In: Mandell GL, Bennett JE, Dolin R,
eds., Principles and Practice of Infectious Diseases, Vol 11. New York: Churchill Livingstone,
1995; 1909-25.
Harman N. Staying the plague. London, Methuen 1917.
Hawes SE, Hillier SL, Benedetti J et al. Hydrogen peroxide-producing lactobacilli and
acquisition of vaginal infections. J Infect Dis 1996; 174:1058-63.
Hillier S, Holmes KK. Bacterial vaginosis. In: Adimom AA, Hamilton H, Holmes KK, Sparling
PF, eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990; 547-59.
Hira SK. Guidelines for Prevention of adverse Outcomes of Pregnancy Due to Syphilis.
Geneva, Switzerland: World Health Organization 1991. Report VDT/91.455.
Hira SK, Bhat GJ, Chikamata DM et al. Syphilis intervention project in pregnancy: Zambian
demonstration project. Genitourin Med 1990; 66:159-64.
Hitchcock P, Fransen L. Preventing HIV infection: lessons from Mwanza and Rakai. Lancet
1999; 353:513-15.
How JH, Bowditch JD. Syphilis in pregnancy: experience from a rural aboriginal community.
Aust N Z J Gynaecol. 1994; 34:383-389.
Hunter D, Maggwa N. Mati G, Tukei M. Mbugua S. Sexual behavior, sexually transmitted
diseases, male circumcision, and risk of HIV infection among women in Nairobi, Kenya.
AIDS 1994; 8:93-99.
Ickovics J, Yoshikawa H. Preventive interventions to reduce hetero-sexual HIV risk for
women: current perspectives, future directions. AIDS 1998; 12(suppl A):SI97-S208.
Irwin K, Mibandumba N, Mbuyi K, Ryder R, Sequeira D. More on vaginal inflammation in
Africa [letter]. N Engl J Med 1993; 328:888-9.
References 115

Jackson DJ, Rakwar JP, Chohan B et al. Urethral infection in a workplace population of East
African men: evaluation of strategies for screening and management. J Infect Dis 1997;
175:833-8.
Jackson D, Ngugi E, Plummer F et al. Stable antenatal HIV-1 seroprevalence with high
population mobility and marked seroprevalence variation among sentinel sites within Nairobi,
Kenya. AIDS 1999; 13:583-9.
Jenniskens F, Obwaka E, Kirisuah S et al. Syphilis control in pregnancy: decentralisation of
screening facilities to primary care level, a demonstration project in Nairobi, Kenya. Int J
Gynecol Obstet 1995; 48(Suppl 1):S121-8.
Jenniskens F, Obwaka E. Syphilis control in pregnant women: clinic based screening,
treatment and counseling services in Nairobi, Kenya. The MotherCare Project. Working
paper:20. 1994.
Joesoef MR, Sumampouw H, Linnan M, Schmid S, Idajadi A, St Louis ME. Douching and
sexually transmitted diseases in pregnant women in Surabaya, Indonesia. Am J Obstet
Gynecol 1996; 174:115-9.
Johansson E, Long NH, Diwan VK, Winkvist A. Gender and tuberculosis control:
perspectives on health seeking behaviour among men and women in Vietnam. Health Policy
2000; 52:33-51.
Johnson AM, Petherick A, Davidson SJ et al. Transmission of HIV to heterosexual partners
of infected men and women. AIDS 1989; 3:367-72.
Joyanes P, Borobio M, Arquez J, Perea E. The association of false-positive rapid plasma
reagin results and HIV infection. Sex Transm Dis 1998; 25:569-571.
Kapiga SH, Lyamuya EF, Lwihula GK, Hunter DJ. The incidence of HIV infection among
women using family planning methods in Dar es Salaam, Tanzania. AIDS 1998; 12:75-84.
Kapiga S, Shao J, Lwihula G, Hunter D. Risk factors for HIV infection among women in Dares-Salaam,
Tanzania. J Acquir Immun Defic Syndr Hum Retrovirol 1994; 7:301-309.
King R, Allen S, Serufilira A, et al. Voluntary confidential HIV testing for couples in Kigali,
Rwanda. AIDS 1993; 7:1393-4.
Kiviat NB, Pavoonen JA, Wolner-Hanssen P, et al. Histopathology of endocervical infection
caused by Chlamydia trachomatis, herpes simplex virus, Trichomonas vaginalis, and
Neisseria gonorrhoeae. Hum Pathol 1990; 21:831-7.
Klein VR, Cox SM, Mitchell MD, et al. The Jarisch-Herxheimer reaction complicating
syphilotherapy in pregnancy. Obster Gynecol 1990; 7S:375-80.
Konde-Lule J, Wawer M, Sewankambo N, et al. Adolescents, sexual behaviour and HIV-1 in
rural Rakai district, Uganda. AIDS 1997; 11:791-799.
Korenromp E, Van Vliet C, Grosskurth H et al. Model based evaluation of single-round mass
treatment of sexually transmitted diseases for HIV control in a rural African population. AIDS
2000; 14:573-93.
Korenromp E, Sudaryo M, De Vlas S et al. What proportion of episodes of gonorrhoea and
chlamydia becomes symptomatic? Int J STD AIDS 2002; 13:91-101.
References 116

Kreiss J, Willeford D, Hensel M, et al. Association between cervical inflammation and
cervical shedding of human immunodeficiency virus DNA. J Infect Dis 1994; 170:1597-1601.
Laga M, Alary M, Nzila Z, et al. Condom promotion, sexually transmitted diseases treatment
and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 1994;
344:246-8.
Laga M, Manoka A, Kivuvu M et al. Non-ulcerative sexually transmitted diseases as risk
factors for HIV-1 transmission in women: results from a cohort study. AIDS 1993; 7:95-102.
Laga M, Nzila N, Goeman J. The interrelationship of sexually transmitted diseases and HIV
infection: implications for the control of both epidemics in Africa. AIDS 1991; 5(suppl):S55-
63.
La Ruche G, Messou N, Ali-Napo L et al. Vaginal douching: association with lower genital
tract infections in African pregnant women. Sex Trans Dis 1999; 26:191-196.
Leroy V, Ladner J, Nyiraziraje M, et al. Effect of HIV-1 on pregnancy outcome in women in
Kigali, Rwanda, 1992-1994. AIDS 1998; 12:643-50.
Leroy V, De Clercq A, Ladner J, Bogaerts J, van de Perre P, Dabis F. Should screening of
genital infections be part of antenatal care in areas of high HIV prevalence? A Prospective
cohort study from Kigali, Rwanda, 1992-1993. Genitourin Med 1995; 71:207-11.
Levine WC, Pope V, Bhoomkar A, et al. Increase in endocervical CD4 lymphocytes among
women with non-ulcerative sexually transmitted diseases. J Infect Dis 1998; 177:167-74
Lilienfeld A, Ulienfeld D. Observational studies: retrospective and cross-sectional studies.
Foundations of epidemiology. 2nd ed. New York: Oxford University Press 1980; 217.
Lilienfeld AM, Graham S. Validity of circumcision status by questionnaire as related to
epidemiological studies of cancer of the cervix. J Natl Cancer Inst 1958; 21:713-720.
Liliestrand J, Bergström S, Nieuwenhuis F, et al Syphilis in pregnant women in Mozambique.
Genitourin Med 1985; 61:355-8.
Lockett AE, Dance DAB, Mabey D, Drasar BS. The use of charcoal supplements in agar for
the growth of Haemophilus ducreyi (abstract E-052). Program and Abstracts of the
Haemophilus ducreyi Symposium, International Society for STD Research, 9th International
Meeting, 6-9 October 1991; 210
Mabey D, Peeling R, Perkins M. Rapid and simple point of care diagnostics for STIs. Sex
Transm Inf 2001; 77:397-8.
Maggwa B, Hunter D, Tukei P, et al. The relationship between HIV infection and cervical
intraepithelial neoplasia among women attending two family planning clinics in Nairobi,
Kenya. AIDS 1993; 7:733-738.
Malonza I, Tyndall MW, Ndinya-Achola J, et al. A randomized, double-blind, placebocontrolled
trial of single-dose ciprofloxacin versus erythromycin for the treatment of chancroid
in Nairobi, Kenya. J Infect Dis 1999; 80:1886-1893.
Mann JM, Nzilambi N, Piot P et al. HIV infection and associated risk factors in female
prostitutes in Kinshasa, Zaire. AIDS 1988; 2:249-54.
References 117

Manning B, Moodley J, Ross SM. Syphilis in pregnant black women. South Afr Med J 1985;
67:966-7.
Mati J, Mbugua S, Ndavi M. Control of cancer of the cervix: feasibility of screening for premalignant
lesions in an African environment. In: Williams P, O'Connor G, Guy B, Johnson C,
eds., Virus-Associated Cancers, 1984.
Mayaud P, Ka-Gina G, Cornelissen J, et al Validation of a \VHO algorithm with risk
assessment for the clinical management of vaginal discharge in Mwanza, Tanzania. Sex
Transm Inf 1998; 74(Suppl 1):S77-84.
Mayaud P, Mosha F, Todd J, et al. Improved treatment services significantly reduce the
prevalence of sexually transmitted diseases in rural Tanzania: results of a randomized
controlled trial. AIDS 1997; 11:1873-80.
Mayaud P, Msuya W, Todd J, et al. STD rapid assessment in Rwandan refugee camps in
Tanzania. Genitourin Med. 1997; 73:33-38.
Mayaud P, Grosskurth H, Changalucha J, et al. Risk assessment and other screening
options for gonorrhoea and chlamydial infections in women attending rural Tanzanian
antenatal clinics. Bull World Health Organ 1995; 73:621-30.
Mazzoli S, Trabattoni D, Lo Caputo S, et al. HIV-specific mucosal and cellular immunity in
HIV-seronegative partners of HIV-seropositive individuals. Nat Med 1997; 3:1250-7
McDermott J, Steketee R, Larsen S, Wirima J. Syphilis-associated perinatal and infant
mortality in rural Malawi. Bull World Health Organ 1993; 71:773-780.
McKenna SL, Muyinda GK, Roth D, et al. Rapid HIV testing and counselling for voluntary
testing centres in Africa. AIDS 1997; 11(suppl 1):S103-10.
Meda N, Sangare L, Lankaonde S, et al. Pattern of sexually transmitted diseases among
pregnant women in Burkina Faso, West Africa: potential for a clinical management based on
simple approaches. Genitourin Med 1997; 73: 188-193.
Meheus A. Sexually transmitted pathogens in mother and newborn. Ann NY Acad Sci 1988;
549:203-13.
Meyer-Weitz A, Reddy P, Van den Borne HW, Kok G, Pietersen J. Health care seeking
behaviour of patients with sexually transmitted diseases: determinants of delay behaviour.
Patient Educ Couns 2000; 41:263-74.
Minkoff HL, Eisenberger-Matityahu D, Feldman J, Burk R, Clarke L. Prevalence and
incidence of gynaecologic disorders among women infected with human immunodeficiency
virus. Am J Obstet Gynecol 1999; 180:824-836.
Morris M, Kretzschmar M. Concurrent partnership and the spread of HIV. AIDS 1997;
11:641-648.
Moses S, Kaul R, Ngugi E et al. A randomised, placebo-controlled trial of monthly
azithromycin to prevent sexually transmitted infections (STI) and HIV in Kenyan female sex
workers (FSWs): preliminary findings. Abstract [ThOrC763] XIII International AIDS
Conference, July 2000, Durban, South Africa.
References 118

Moses S, Bailey RC, Ronald AR. Male circumcision: assessment of health benefits and
risks. Sex Transm Infect 1998; 74:368-373.
Moses S, Ngugi E, Jackson D, et al. Declines in syphilis prevalence in pregnancy and in
STD syndromes among clinic attendees at five primary health care centres in Nairobi, Kenya,
1993 to 1997 (Abstract 22190). 12th World AIDS Conference, Geneva July 1998.
Moses S, Muia E, Bradley JE et al. Sexual behavior in Kenya: implications for sexually
transmitted disease transmission and control. Soc Sci Med 1994; 39:1649-1656.
Moses S, Ngugi EN, Bradley JE et al. Healthcare-seeking behavior related to the
transmission of sexually transmitted diseases in Kenya. Am J Public Health 1994; 84:1947-
1951.
Moses S, Bradley J, Muia E et al. Evaluation of an in-service training program in STI
management for clinical officers and nurses in Kenya. IIX th International Conference on AIDS
in Africa, 1992, Cameroon. [Abstract WP 211]
Moss G, Overbaugh J, Welch M, et al. Human immunodeficiency virus DNA in urethral
secretions in men: association with gonococcal urethritis and CD4 depletion. J Infect Dis
1995; 172:1469-1474.
Nandwani R, Evans D. Are you sure it's syphilis? A review of false positive serology. Int J
STD AIDS 1995; 6:241-248.
National Cancer Institute Workshop. The 1988 Bethesda system for reporting
cervical/vaginal cytological diagnosis. JAMA 1989; 262:9314.
Ndinya-Achola JO, Ghee AE, Kihara AN, et al. High HIV prevalence, low condom use and
gender differences in sexual behaviour among patients with STD-related complaints at a
Nairobi primary health care clinic. Int J STD AIDS 1997; 8:506-514.
Ndoye I, Mboup S, De Schrijver A, et al. Diagnosis of sexually transmitted infections in
female prostitutes in Dakar, Senegal. Sex Transm Inf 1998; 74:SII2-17.
Newell J, Senkoro K, Mosha F, et al. A population based study of syphilis and sexually
transmitted diseases syndromes in northwestern Tanzania. Risk factors and health seeking
behaviour Genitourin Med 1993; 69:421-426.
Ngugi EN, Wilson D, Sebstad J, Plummer FA, Moses S. Focused peer-mediated educational
programs among female sex workers to reduce sexually transmitted disease and human
immunodeficiency virus transmission in Kenya and Zimbabwe. J Infect Dis 1996;
174(suppl2):S240-7.
Ngugi EN, Plummer FA, Simonsen NJ, et al. Prevention of transmission of human
immunodeficiency virus in Africa: effectiveness of condom promotion and health education
among prostitutes. Lancet 1988; 2:887-90.
Njeru E, Eldridge G, Ngugi E, Plummer F, Moses S. STD partner notification and referral in
primary level health centres in Nairobi, Kenya. Sex Transm Dis 1995; 22:231-235.
Nugent RP, Krohn MA, Hillier S. Reliability of diagnosing bacterial vaginosis is improved by a
standardized method of Gram stain interpretation. J Clin Micro 1991; 29:297-301.
References 119

Oberlander L, Elverdam B. Malaria in the United Republic of Tanzania: cultural
considerations and health-seeking behaviour. Bull World Health Organ 2000; 78:1352-7.
Opai-Tetteh ET, Hoosen AA, Moodley J. Re-screening for syphilis at the time of delivery in
areas of high prevalence. South Afr Med J 1993; 83:725-6.
Pedersen Sheller J et al. HIV-infection, syphilis or genital lesions in Maun, Botswana. Ugesk
Laeger 1990, 20 :1441-43.
Peterman TA. Facilitators of HIV transmission during sexual contact. In Heterosexual
Transmission of AIDS. Edited by Alexander NJ, Gabelnick HL, Spieler JM. New York: Alan
R. Liss; 1990.
Pido D, Jenniskens E. MotherCare Maternal Congenital Syphilis Control Project Qualitative
Research Report. Arlington, Va: MotherCare/ John Snow Inc; 1993. MotherCare working
paper 18.
Piot P, Tezzo R. The epidemiology of HIV and other sexually transmitted infections in the
developing world. Scand J Infect Dis 1990; 69:89-97.
Piot P & Holmes K.K. Sexually transmitted diseases. In Tropical and Geographical Medicine,
New York: McGraw-Hill, 1984; 844-862.
Plummer F, Simonsen NJ, Cameron D, et al. Cofactors in male-female sexual transmission
of human immunodeficiency virus type 1. J Infect Dis 1991; 163:233-9.
Potterat JJ, Meheus A, Gallwey J. Partner notification: operational considerations. Int J STD
AIDS 1991; 2:411-5.
Qolohle DC, Hoosen AA, Moodley J, Smith AN, Milsana KP. Serological screening for
sexually transmitted infections in pregnancy: is there any value in re-screening for HIV and
syphilis at the time of delivery? Genitourin Med 1995; 71:65-7.
Quigley M, Weiss H, Hayes R. Male circumcision as a measure to control HIV infection and
the other sexually transmitted diseases. Current Opinion in Infectious Diseases 2001; 14:71-
75
Quigley M, Munguti K, Grosskurth H et al. Sexual behaviour patterns and other risk factors
for HIV infection in rural Tanzania: a case-control study. AIDS 1997; 11:237-248.
.
Rajamanoharan S, Low N, Jones S, Pozniak A. Bacterial vaginosis, ethnicity, and the use of
genital cleaning agents: a case-control study. Sex Transm Dis 1999; 404-9.
Rakwar J, Kidula N, Fonck K, Kirui P, Ndinya-Achola J, Temmerman M. HIV/STD: the
women to blame? Knowledge and attitudes among STD clinic attendees in the second
decade of HIV/AIDS. Int J STD AIDS 1999; 10:543-547.
Ramjee G, Abdool Karim S, Sturm A. Sexually transmitted infections among sex workers in
Kwa Zulu-Natal, South Africa. Sex Transm Dis 1998; 25:346-9.
Rao P, Mohamedali FY, Temmerman M, Fransen L. Systematic analysis of STD control: an
operational model. Sex Transm Inf 1998; 74(suppl 1):S17-S22.
Ratnam AV, Din SN, Hira SK, et al. Syphilis in pregnant women in Zambia. Br J Vener Dis
1982; 58:355-8.
References 120

Resources for Child Health Project (REACH): Nairobi Area Study. Nairobi, Kenya: USAID,
1988.
Ridgway GL, Young LS, eds. Azithromycin in the management of STDs and opportunistic
infections. Int J STD AIDS 1996; 7(suppl1):1-12.
Rotchford K, Lombard C, Zuma K, Wilkinson D. Impact on perinatal mortality of missed
opportunities to treat maternal syphilis in rural South Africa: baseline results from a clinic
randomized controlled trial. Trop Med Int Health 2000; 5:800-4.
Runganga A, Pitts M, McMaster J. The use of herbal and other agents to enhance sexual
experience. Soc Sci Med 1992; 35:1037-42.
Rutgers S. Syphilis in pregnancy: a medical audit in a rural district. Cent Afr J Med. 1993;
39(12):248-253.
Ryan CA, Courtois BN, Hawes SE et al. Risk assessment, symptoms, and signs as
predictors of vulvovaginal and cervical infections in an urban US STD clinic: implications for
use of STD algorithms. Sex Transm Inf 1998; 74(suppl 1):59-76.
Ryder R, Temmerman M. The effect of HIV-1 infection during pregnancy and the perinatal
period on maternal and child health in Africa. AIDS 1991; 5(suppl 1):S75-S85.
Ryder RW, Ndilu M, Hassig SE et al. Heterosexual transmission of HIV-1 among employees
and their spouses at two large businesses in Zaire. AIDS 1990; 4:725-32.
Sangare L, Meda N, Lankoande S, et al. HIV infection among pregnant women in Burkina
Faso; a nation wide survey. Int J ST'D AIDS 1997; 8(10):646-51.
Schacker T, Ryncarz A, Goddard D, et al. Frequent recovery of HIV-1 from genital herpes
simplex virus lesions in HIV-1 infected men. JAMA 1998; 280:61-66.
Schmid G, Markowitz L, Joesoef R, Koumans E. Bacterial vaginosis and HIV infection
[editorial]. Sex Transm Inf 2000; 76:3-4.
Schmid G. Serologic screening for syphilis. Rationale, cost, and realpolitik. Sex Transm Dis
1996; 23:45-50.
Scholes D, Stergachis A, Ichikawa LE, Heidrich FE, Holmes KK, Stamm WE. Vaginal
douching as a risk factor for cervical Chlamydia trachomatis infection. Obstet Gynecol 1998;
91:993-7.
Scholes D, Darling JR, Stergachis A, Weiss NS, Wang SP, Grayston JT. Vaginal douching
as a risk factor for acute pelvic inflammatory disease. Obstet Gynecol 1993; 81:601-6.
Schultz K, Murphy F, Patamasucon P, Meheus A. Congenital syphilis. In: Holmes K, March
P, Sparling P, et al., eds. Sexually Transmitted Diseases. 2nd ed. New York, NY. McGraw-
Hill; 1990; 821-840.
Schultz KF, Cates WJr, O'Mara PR. Pregnancy loss, infant death, and suffering: legacy of
syphilis and gonorrhoea in Africa. Genitourin Med 1987; 63:320-5.
References 121

Serwadda D, Gray R, Wawer M et al. HIV-1 incidence and prevalence among pregnant
women in a population based rural cohort, Rakai district, Uganda. Vancouver: XI
International Conference on AIDS, 1996.
Serwadda D, Gray RH, Wawer MJ et al. The social dynamics of HIV transmission as
reflected through discordant couples in rural Uganda. AIDS 1995; 9:745-50.
Sewankambo N, Gray R, Wawer M, et al. HIV-1 infection associated with abnormal vaginal
flora morphology and bacterial vaginosis. Lancet 1997; 350:546-50.
Simonsen NJ, Plummer FA, Ngugi EN, et al. HIV infection among lower socio-economic
strata prostitutes in Nairobi. AIDS 1990; 4:139-44.
Steen R, Soliman C, Bucyana S, Dallabetta G. Partner referral as a component of integrated
sexually transmitted disease services in tow Rwandan towns. Genitourin Med 1996; 72:56-9.
Stray-Pedersen B. Cost-benefit analysis of a prenatal preventive program against congenital
syphilis. NIPH Ann. 1980; 3:57-66.
Taha T, Hoover D, Dallabetta G, et al. Bacterial vaginosis and disturbances of vaginal flora:
association with increased acquisition of HIV. AIDS 1998; 12:1699-706.
Temmerman M, Gichangi P, Fonck K et al. Effect of a syphilis control program on pregnancy
outcome in Nairobi, Kenya. Sex Transm Inf 2000; 76:117-21.
Temmerman M, Fonck K, Bashir F, et al. Declining syphilis prevalence in pregnant women in
Nairobi since 1995: another success story in the STD field? Int J STD AIDS 1999; 10: 405-
408.
Temmerman M. Final report: AIDS/STD control programme in the Republic of Kenya.
European Commission, 1999.
Temmerman M, Tyndall MW, Kidula N, Claeys P, Muchiri L, Quint W. Risk factors for human
papillomavirus and cervical precancerous lesions, and the role of concurrent HIV-1 infection.
Int J Obstet Gynecol 1999; 65:171-81.
Temmerman M, Kidula N, Tyndall MW, Rukuria R. Muchiri L, Ndinya-Achola JO. The
supermarket for women's reproductive health: the burden of genital infections in a family
planning clinic in Nairobi, Kenya. Sex Transm Infect 1998; 74:202-204.
Temmerman M, Ndinya-Achola J, Ambani J, Piot P. The right not to know HIV-test results.
Lancet. 1995; 345:969-970.
Temmerman M, Njagi E, Nagelkerke N, Ndinya-Achola J, Plummer F, Meheus A. Mass
antimicrobial treatment in pregnancy. J Reprod Med. 1995; 40:176-180.
Temmerman M, Chomba E, Ndinya-Achola J, Plummer F, Coppens M, Piot P. Maternal HIV-
1 infection and pregnancy outcome. Obstet Gynecol 1994; 7:20-4.
Temmerman M, Mohamed Ali F, Fransen L. Syphilis prevention in pregnancy: an opportunity
to improve reproductive and child health in Kenya. Health Policy and Planning 1993; 8:122-7.
Temmerman M, Lopita M, Sanghvi H, Sinei SK, Plummer FA, Piot P. The role of maternal
syphilis, gonorrhoea and HIV-1 infections in spontaneous abortion. Int J STD AIDS. 1992; 3:
418-22.
References 122

Temmerman M, Ali FM, Ndinya-Achola J, Moses S, Plummer FA, Piot P. Rapid increase of
both HIV-l infection and syphilis among pregnant women in Nairobi, Kenya. AIDS 1992;
6:1181-5.
Temmerman M, Lopita M, Sinei S, Plummer FA, Nagelkerke N, Piot P. Sexually transmitted
infections as risk factors for spontaneous abortion. Int J STD AIDS 1992; 3.418-22.
Temmerman M, Jenniskens F et al. Prevalence rates of STDs in women at the STD referral
clinic and at a MCH clinic in Nairobi, Kenya. VII International Conference on AIDS in Africa,
1996, Kampala, Uganda.
Temmerman M, Mohamedali F, Ndinya-Achola J, et al. Rapid increase of both HIV-1
infection and syphilis among pregnant women in Nairobi, Kenya. AIDS 1992; 6:1181-1185.
Temmerman M, Plummer F, Farah A, et al. Gonorrhoea in pregnancy. J Obstet Gynecol
1992; 12:162-6.
Temmerman M, Lopita MI, Sanghvi HC, Sinei SK, Plummer FA, Piot P. The role of maternal
syphilis, gonorrhoea and HIV-1 infections in spontaneous abortion. Int J STD AIDS 1992;
3:418-22.
Temmerman M, Plummer F, Mirza N, et al. Infection with human immunodeficiency virus
(HIV) as a risk factor for adverse obstetrical outcome. AIDS 1990; 4:1087-1093.
Thomas T, Choudri S, Kariuki C, et al. Identifying cervical infection among pregnant women
in Nairobi, Kenya: limitations of risk assessment and symptom based approaches. Genitourin
Med 1996; 2:334-8.
Tyndall MW, Agoki E. Malisa W, Ndinya-Achola JO, Ronald A, Plummer FA. Sexual behavior
and perceived risk of AIDS among men in Kenya attending a clinic for sexually transmitted
disease. Clin Infect Dis 1994; 19:441-447.
Tyndall MW, Kidula N, Sande J, Ombette J, Temmerman M. Predicting Neisseria
gonorrhoeae and Chlamydia trachomatis infection using risk scores, physical examination,
microscopy, and leukocyte esterase dipsticks among asymptomatic women attending a
family planning clinic in Kenya. Sex Transm Dis 1999; 26:476-82.
Tyndall MW, Omari S, Malonza I et al. High prevalence of urethral infection among men with
non-specific symptoms at an STI clinic in Kenya. XII th World AIDS Conference, Geneva,
1998 [abstract 33242].
Ulin P. African women and AIDS: negotiating behavioural change. Soc Sci Med 1992; 34:63-
73.
UNAIDS. Epidemiological fact sheet on HIVIAIDS and sexually transmitted diseases.
Geneva: World Health Organisation, 1998.
Urassa M, Todd J, Ties Boerma J, Hayes R, Isingo R. Male circumcision and susceptibility to
HIV infection among men in Tanzania. AIDS 1997; 11:73-80.
US Bureau of the Census HIV/AIDS Surveillance Data Base 1997.
US Census bureau 2000, website (www.census.gov/cgi-bin/icp).
References 123

Van Dam J, Anastasi MC. Male Circumcision and HIV prevention: directions for future
research. Report Population Council, 2000.
Van Dam J, Dallabetta G, Piot P. Prevention and control of sexually transmitted diseases in
developing countries. In Holmes K, Sparling PF, Mardh P, et al. eds., Sexually Transmitted
Diseases, USA, 1999.
Van De Wijgert JH, Mason PR, Gwanzura L et al. Intravaginal practices, vaginal flora
disturbances and acquisition of sexually transmitted diseases in Zimbabwean women. J
Infect Dis 2000 ; 181:587-94.
Van Vliet C, Meester E, Korenromp E, Singer B, Bakker R, Habbema JD. Focusing
strategies of condom use against HIV in different behavioral settings: an evaluation based on
a simulation model. Bull WHO 2001; 79:442-454.
Voeten HA, Otido JM, O’Hara HB et al. Quality of sexually transmitted diseases case
management in Nairobi, Kenya: a comparison among different types of healthcare facilities.
Sex Transm Dis 2001; 28:633-42.
Vuylsteke B, Lap M, Alary M, et al. Clinical algorithms for the screening of women for
gonococcal and chlamydial infection: evaluation of pregnant women and prostitutes in Zaire.
Clin Infect Dis 1993; 17:82-8.
Wasserheit J. Epidemiological synergy: interrelationships between human immunodeficiency
virus infection and other sexually transmitted diseases. Sex Transm Dis 1992; 19:61-77.
Watts TE, Larsen SA, Brown ST. A case-control study of stillbirths at a teaching hospital in
Zambia, 1979-1980: serological investigations for selected infectious agents. Bull WHO
1984, 62:803-8.
Wawer M, Sewankambo N, Serwadda D, et al. Control of sexually transmitted diseases for
AIDS prevention in Uganda: a randomized community trial. Lancet 1999; 353:525-35.
Wawer M, Gray R, Sewankambo N, et al. A randomized, community trial of intensive
sexually transmitted disease control for AIDS prevention, Rakai, Uganda. AIDS 1998;
12:1211-25.
Weiss HA, Quigley MA, Hayes RJ. Male circumcision and risk of HIV infection in sub-
Saharan Africa: a systematic review and meta-analysis. AIDS 2000; 14:2361-70.
WHO/GPA. Informal technical working group meeting on STD activities in GPA. The
evaluation of algorithms for the diagnosis and treatment of vaginal discharge. Agenda item
No IV, background paper No 5. Geneva: WHO, 1993.
WHO-UNAIDS. Sexually Transmitted Diseases: policies and principles for prevention and
care. UNAIDS/97.6.
WHO. Integration of STI management into family planning services: what are the benefits?
www. who.int/RHR 1999.
WHO. Global prevalence and incidence of selected curable sexually transmitted infections:
overview and estimates. WHO/CDS/CSR/EDC/2001.10.
References 124

Wilkinson D, Ndovela N, Harrison A, et al. Family planning services in developing countries:
an opportunity to treat asymptomatic and unrecognised genital tract infections? Genitourin
Med. 1997; 73:558-560.
Wilkinson D, Sach M, Connolly C. Epidemiology of syphilis in pregnancy in rural South
Africa: opportunities for control. Trop Med Int Health 1997; 2:57-62.
Williams AO. More on vaginal inflammation in Africa [letter]. N Engl J Med 1993; 328:888.
Winfield J, Latif AS. Tracing contacts of persons with sexually transmitted diseases in a
developing country. Sex Transm Dis 1985; 12:5-7.
Wölner-Hanssen P, Eschenbach DA, Paavonen J et al. Association between vaginal
douching and acute pelvic inflammatory disease. JAMA 1990; 263:1936-41.
World Bank. Investing in Health. New York, NY. Oxford University Press 1993;115.
Yala F et al. Enquête virologique et bactériologique sur l’infection materno-foetale à
Brazzaville. Bull Soc Path Ex 1991; 84:627-34.
Zhang J, Thomas G, Leybovich E. Vaginal douching and adverse health effects: a metaanalysis.
Am J Public Health 1997; 87:1207-11.
References 125

WHO Collaborating Center on HIV/AIDS, Nairobi, Kenya
A NNEX
The WHO Collaborating Group at the University of Nairobi has played a major role in STI/HIV
research in Kenya and has resulted in world-renowned data. The group started as a
partnership between the University of Nairobi, the Nairobi City Council, the University of
Manitoba (Winnipeg, Canada), the Institute of Tropical Medicine (Antwerp, Belgium) and the
University of Washington (Seattle, USA). The group has become larger over the years as the
good working environment, the optimal research infrastructures, and the collaborative
atmosphere attracted other universities. During the study period, the group counted in
addition the Oxford University (Oxford, England), the Erasmus University (Rotterdam,
Holland) and the Ghent University (Ghent, Belgium). The Institute of Tropical Medicine was
no longer part of the group.
The shared interest in STI, especially genital ulcer disease (GUD) of the Department Heads
of Medical Microbiology at the University of Nairobi and the University of Manitoba, Herbert
Nsanze and Allan Ronald respectively, created the opportunity that sparked the collaboration
in 1980. Two senior infectious disease fellows, Margaret Fast during 1980 and Frank
Plummer during 1981 began the Nairobi-Manitoba exchange. Laboratory space was
provided at the University of Nairobi, Department of Medical Microbiology. Dr DaCosta
invited the team to begin clinical research in the special Treatment Clinic, a referral clinic for
STI in downtown Nairobi.
In 1981, Dr Peter Piot from the Institute of Tropical Medicine in Antwerp joined the
partnership. Over the subsequent decade, three Belgian physicians, Lieve Fransen, Marie
Laga and Marleen Temmerman worked in Nairobi, focusing on STI/HIV in reproduction.
Marleen Temmerman, after becoming Professor in obstetrics-gynecology at the Ghent
University, and where she created the International Center for Reproductive Health (ICRH)
continued the collaboration but now out of this institution and this from 1992.
In 1983, the World Health Organization (WHO) formally recognized the research group and
designated it as a WHO Collaborating Center for Sexually Transmitted Diseases.
Annex 126

In 1982, Dr Elizabeth Ngugi, a senior nursing- sister in the Ministry of Health, initiated a
community self-improvement program among commercial sex workers in the Pumwani area
of Nairobi and opened a second front for ongoing research into the control of STI.
In 1985, the group was joined by the department of Dr King Holmes from the University of
Washington. The Seattle group identified the spread of HIV among the prostitute cohort and
a series of studies were implemented to understand the heterosexual epidemiology of the
virus and the risk factors associated with its rapid spread.
From 1986 onwards, the research was almost entirely devoted to the epidemiology of HIV
and AIDS, and its parallelism with the STI already being investigated.
In 1992, Dr Joan Kreiss from the University of Washington set up another research site in
Mombasa, on the Kenyan coast, starting up a cohort of female sex workers. The site chosen
was the Ganjoni clinic, an infectious disease clinic where female sex workers are regularly
seen for medical check-up. In 1993, the Ghent University joined in with the EU program
mainly focusing on STI prevention and control as well as operational research. The Ghent
University also did research on mother-to-child transmission of HIV at the provincial hospital
in Mombasa.
The Erasmus University started its collaboration with the group in 1996. The main objective
of their work was the testing of a computer modeling to predict the spread of HIV in a given
population. As the previous studies from the other collaborating universities had generated
an amount of invaluable data on several key issues, the Nairobi site was an ideal setting to
test the model.
The Oxford University also joined the Nairobi group. The University had been working on an
AIDS vaccine, based on earlier studies from the collaborating group. The site was prepared
to start the first human HIV trial in Africa.
During the entire study period, Dr Job Bwayo has been the Head of the Department of
Medical Microbiology. Professor Jeconiah Ndinya-Achola, the former Head of Department,
continued to play a major role in the research activities of the center.
Annex 127

S UMMARY
Sexually transmitted infections (STIs) have, in all likelihood, always been present. The recent
emergence of HIV/AIDS has however added new relevance to STIs and has also reinforced
the tremendous cultural burden of diseases associated with sexual contact.
This work focuses on the epidemiology of STI among different population groups in Nairobi,
Kenya. We studied women at antenatal clinics as well as at maternity level, men and women
seeking care for STI complaints, men and women seeking general health care, and female
sex workers.
A cross sectional study of 520 female patients in the referral STI clinic took place to
determine the etiologies of STIs and presence of cervical dysplasia (CIN). The prevalence
rates were: HIV 29%, candidiasis 35%, trichomoniasis 25%, bacterial vaginosis 16%,
gonorrhoea 6%, chlamydia 4%, syphilis seropositivity 6%, genital warts 6% and genital
ulcers 12%. CIN was found in 12%. Factors related to sexual behavior, especially the
number of sex partners, were associated with several STI. Gonorrhoea, bacterial vaginosis,
CIN and genital warts/ulcers were independently associated with HIV infection. These data
indicate an alarming prevalence of STI/HIV, especially in the young age groups, and the lack
of safe sex methods, and call for more aggressive prevention campaigns. Risk factors for
STI were mainly related to the number of sex partners.
Sexual behavior and health-seeking behavior are important components of STI control. A
study to assess this behavior was done among 471 patients attending the STI referral clinic,
and among 555 patients attending three PHC clinics in low socio-economic areas. It was
found that a large proportion of patients had sought treatment in public and private sectors
before attending the STI clinic. Major gender differences in delay of health-seeking for STI
were observed with women waiting longer than men to seek medical care. In addition,
women more than men had sex while symptomatic, mostly with their regular partner.
Condoms were rarely used during illness. Men at the STI clinic admitted extra-marital affairs
in 68% and paying for sex in 30%, but blamed their wives for their STI. At the PHC clinics,
men reported extra-marital affairs in 17% versus 8% for women. Women’s knowledge about
health in general and about STI in particular was poor. Significantly more men than women
reported a history of STI. A high prevalence of gonorrhoea (3%) and chlamydia (6%) was
found among the PHC patients with no difference between sexes.
Summary 128

We evaluated the validity of different algorithms for the diagnosis of gonococcal and
chlamydial infections among 621 pregnant and non-pregnant women consulting for
complaints of vaginal discharge in 3 City Council clinics. The mean age was 24 years and
334 (54%) were pregnant. The overall prevalence rates were: 50% candidiasis, 23%
trichomoniasis, 9% bacterial vaginosis, 7% gonorrhoea, 9% chlamydia, 7% syphilis and 22%
HIV. In non pregnant women, gonococcal and chlamydial infection was significantly
associated with 1) demographic and behavioral risk markers such as being single, younger
than 20 years, multiple sex partners in the previous 3 months; 2) symptom fever; and 3)
signs including presence of yellow or bloody vaginal discharge, cervical mucopus, cervical
erythema and friability. Among pregnant women only young age, dysuria and fever were
significantly associated with cervical infection. However, none of these variables on itself was
both sensitive and specific enough for the diagnosis of cervical infection. Several algorithms
were generated and applied to the study data. The algorithm including risk markers
performed slightly better than the current Kenyan algorithm. We concluded that STIs form a
major problem in the Nairobi area and should be addressed accordingly. None of the tested
algorithms for the treatment of vaginal discharge would constitute a marked improvement of
the existing flowchart. Hence better detection tools for the specific etiology of vaginal
discharge are urgently needed.
An analysis of data collected through routine program monitoring indicates that syphilis
serology in pregnant women has declined in Nairobi from 7.3% to 3.8% between 1993 and
1997. We analyzed routinely collected data of 1998 to describe the performance and costs of
a decentralized syphilis-screening program in pregnant women. RPR seroprevalence was
3.4%. The program succeeded in screening and treating women and partners properly, but
failed in performing reliable RPR tests. The cost per averted case was calculated between
$95 and $112. The sustainability of this labor-intensive program is threatened by costs and
logistic constraints. Alternative strategies, such as mass epidemiological treatment of
pregnant women should be considered.
As part of a larger study on neonatal outcome of syphilis during pregnancy, we examined the
determinants of partner notification among pregnant syphilitic women. Of the 12,414 women
tested for syphilis at delivery, 377 (3%) were RPR reactive. Only 53% of those women
informed their partner and 36% of the partners were treated. Adverse pregnancy outcome
was related to lack of treatment of the partner during pregnancy (2% versus 15%, OR 6.1,
95% CI 0.8-45). This study shows lower partner notification/treatment rates in syphilisinfected
women identified at delivery than during pregnancy. These findings suggest that
messages focusing on the health of the unborn child have a positive effect on partner
Summary 129

notification. Innovative and locally adapted strategies for partner notification are needed in
the fight against HIV/STI.
A placebo controlled trial of regular antibiotic prophylaxis with azithromycin to prevent STIs
and HIV in a cohort of female sex workers (FSW) was initiated in 1998. We described the
baseline characteristics of 318 women screened for study enrolment. The mean age was 32
years and mean duration of sex work 7 years. The average number of clients per day was 4.
High-risk behavior was frequent: 9% practiced anal intercourse, 19% sex during menses,
and 3% used intravenous drugs. While 20% reported condoms use with all clients, 37%
never use condoms. Despite this high-risk behavior, STI and HIV frequency was relatively
low: HIV-1 27%, bacterial vaginosis 46%, trichomoniasis 13%, gonorrhoea 8%, chlamydia
7% and syphilis 6%. CIN was infrequent (3%) and tended to be more common in HIV
infected FSWs. Vaginal douching was highly prevalent in this population and was
significantly associated with bacterial vaginosis. In this study we have identified a large
population of HIV uninfected, but high-risk, FSW. It appears feasible to access these women
with prevention programs, including a proposed trial of HIV prevention through STI
chemoprophylaxis.
We concluded from these different studies that STI in Nairobi have already spread from high
risk groups to the general population with men playing an important role as bridging group
between FSWs and their spouses/regular partners. Interventions specifically targeted to men
to enhance a more responsible attitude towards safer sex, while maintaining interventions
directed to the high risk groups, seem necessary.
We can also deduct from our results that a large proportion of people infected with STIs are
not timely nor correctly treated. Several reasons appeared to be causing this: 1) health care
for STIs is sought late, especially among women, and is sought from non-trained providers;
and 2) the algorithm used for diagnosis of vaginal discharge is not performing very well.
Innovative strategies to address these issues are urgently needed.
We showed that the cost of the syphilis-screening programme in Nairobi is high while not
being very effective in case finding. Presumptive treatment of all pregnant women and their
partners could be considered. Mass treatment of FSWs was shown to be feasible.
Summary 130

R ESUME
Les Infections Sexuellement Transmissibles (IST) ont très probablement toujours existé.
L’émergence récente du VIH/SIDA a ajouté une nouvelle dimension aux IST et a aussi
renforcé le fardeau culturel énorme des maladies associées au contact sexuel.
Ce travail se focalise sur l’épidémiologie des IST parmi différentes populations cibles à
Nairobi, Kenya. Nous avons étudié des femmes dans les cliniques prénatales et à la
maternité, des hommes et des femmes qui fréquentaient des cliniques pour des problèmes
d’IST, des hommes et des femmes dans des consultations générales, et des travailleuses du
sexe.
Une étude transversale de 520 patientes dans la clinique de référence des IST a eu lieu afin
de déterminer les étiologies des IST et la présence des dysplasies cervicales. Les
prévalences étaient : VIH 29%, candidiasis 35%, trichomoniasis 25%, vaginosis bactérienne
16%, gonorrhée 6%, chlamydia 4%, syphilis 6%, verrues génitaux 6% et ulcères génitaux
12%. NIC était présent en 12%. Des facteurs relatifs au comportement sexuel, surtout le
nombre des partenaires sexuels, étaient associés avec plusieurs IST. Gonorrhée, vaginosis
bactérienne, NIC et verrues/ulcères génitaux étaient indépendamment associés à l’infection
du VIH. Ces données indiquent une prévalence alarmante des IST/VIH, particulièrement
parmi les jeunes groupes d’âge, l’absence de comportements sexuels sans risque, et fait
appel à des campagnes de prévention plus agressive. Les facteurs de risque des IST étaient
surtout liés au nombre des partenaires sexuels.
Le comportement sexuel et le comportement de « health-seeking » sont des volets
importants pour le contrôle des IST. Nous avons étudié ce comportement parmi 471 patients
à la clinique de référence des IST et parmi 555 patients à 3 cliniques dans des endroits à
bas niveau socio-économique. On a découvert qu’une proportion importante des patients
avait cherché un traitement dans le secteur public et privé avant de venir à la clinique IST.
Des différences importantes de genre par rapport au délai de chercher l’aide médicale ont
été observées. Beaucoup plus de femmes que d’hommes avaient des rapports sexuels
pendant la phase symptomatique, et en général avec leur partenaire régulier. Des condoms
sont rarement utilisés pendant la maladie. 68% des hommes qui fréquentent la clinique IST
avouent des relations hors mariages et 30% paient en échange pour le sexe, mais pourtant
ils accusent leurs femmes quand ils sont atteints d’IST. Au niveau des cliniques primaires,
17% des hommes mentionnent des relations hors mariage contre 8% de femmes. La
Résumé 131

connaissance des femmes sur la santé en général et les IST en particulier, est faible.
Significativement plus d’hommes que de femmes rapportent une histoire d’IST. Une
prévalence élevé de gonorrhée (3%) et de chlamydia (6%) a été trouvée parmi les patients
des cliniques primaires, sans différence notoire entre les sexes.
Nous avons évalué la validité des différents algorithmes pour le diagnostic des infections de
gonorrhée et chlamydia parmi 621 femmes qui consultaient des services médicaux pour
écoulement vaginal dans 3 cliniques. L’âge moyen était 24 ans et 334 (54%) étaient
enceintes. Les taux de prévalence étaient: candidiasis 50%, trichomoniasis 23%, vaginosis
bactérienne 9%, gonorrhée 7%, chlamydia 9%, syphilis 7% et VIH 22%. Parmi les femmes
non enceintes, l’infection de gonorrhée ou de chlamydia était significativement associée avec
1) facteurs de risques démographique et comportemental comme être célibataire, avoir
moins de 20 ans, plusieurs partenaires sexuels durant les 3 mois avant l’enquête; 2) le
symptôme de fièvre; et 3) des signes comme la présence d’un écoulement vaginal purulent
ou saignant, mucopus cervical, érythème cervical et friabilité. Parmi les femmes enceintes,
seuls jeune âge, la dysurie et la fièvre étaient significativement associés avec une infection
cervicale. Pourtant, aucunes de ces variables était en même temps assez sensitive et
spécifique pour le diagnostic de l’infection cervicale. Plusieurs algorithmes ont été générés et
appliqués aux données. L’algorithme incluant des marqueurs de risqué avait une
performance un peu mieux que celui actuellement en usage au Kenya. Nous avons conclu
que les ISTs représentent un problème majeur et devrait être prise en charge de façon
adéquate. Aucun des algorithmes testés pour le traitement de l’écoulement vaginal ne
constitue une amélioration significative d’algorithme existant. Il y a une urgence à développer
des méthodes de détection pour les étiologies spécifiques de l’écoulement vaginal.
Une analyse des données collectées à travers le programme de routine de suivi indique que
la sérologie de la syphilis parmi les femmes enceintes a diminué de 7,3% à 3,8% entre 1993
et 1997. Nous avons fait une analyse des données de 1998 afin de décrire la performance et
le coût du programme décentralisé de dépistage de la syphilis parmi les femmes enceintes.
La séroprévalence du RPR était de 3,4%. Le programme réussissait à traiter les femmes et
leurs partenaires, mais échouait à réaliser des tests RPR fiables. Le coût par cas évité était
calculé entre 95 et 112 USD. La pérennité de ce programme est contrariée par le coût et les
besoins en logistique. Des stratégies alternatives comme le traitement épidémiologique des
femmes enceintes devrait être considéré.
Faisant partie de l’étude sur l’effet congénital de syphilis, nous avons examiné les
déterminants de la notification du partenaire. Des 12,414 femmes testées pour syphilis à
Résumé 132

l’accouchement, 377 (3%) étaient RPR réactives. Seules 53% de ces femmes informaient
leur partenaire et 36% des partenaires étaient traités. Un effet néonatal négatif était lié à un
manque de traitement du partenaire pendant la grossesse (2% versus 15%, OR 6,1, 95% CI
0,8-45). Cette étude montre que les taux de notification et traitement des partenaires est plus
bas après la naissance que pendant la grossesse. Ceci indique que les messages qui
mettent l’accent sur la santé du fœtus ont un effet positif sur la notification. Des stratégies
innovatrices et adaptées localement pour une meilleure notification du partenaire sont
nécessaires dans la lutte contre les IST/VIH.
Une étude de prophylaxie régulière avec azithromycin pour prévenir des IST/VIH dans une
cohorte des travailleuses du sexe était initiée en 1998. 318 femmes ont été testées pour
enroulement. L’âge moyen était de 32 ans et la durée moyenne de travail du sexe de 7 ans.
Le nombre moyen de clients par jour était de 4. Des comportements à hauts risques étaient
fréquents: 9% pratiquaient des rapports sexuels par voie anale, 19% pendant la
menstruation et 3% utilisaient des drogues parentérales. Tandis que 20% disait d’utiliser des
condoms avec tous les clients, 37% n’en utilisaient jamais. Malgré ces comportements à
hauts risques, le taux des IST et VIH était relativement bas: VIH 27%, vaginosis bactérienne
46%, trichomoniasis 13%, gonorrhée 8%, chlamydia 7% et syphilis 6%. NIC n’était pas
fréquent (3%) et avait tendance à être plus commun parmi les femmes porteuses de VIH. La
pratique de la douche vaginale est très prévalent dans cette population et est
significativement associée avec la vaginosis bactérienne. Il paraît faisable d’atteindre les
travailleuses du sexe avec des programmes de prévention, y compris une étude de
prévention du VIH à travers la chimioprophylaxie des IST.
Nous pouvons conclure à l’issue de ces différentes études que les IST sont déjà diffusés du
groupe à hauts risques vers la population générale avec les hommes jouant un rôle
important comme « bridging » group entre les travailleuses du sexe et leurs partenaires. Des
interventions spécifiquement ciblées aux hommes dans le but d’augmenter une attitude plus
responsable vis-à-vis des rapports sexuels à moindres risques semblent nécessaires, tout en
continuant des interventions dirigées vers des groupes à hauts risques.
Nous pouvons également déduire de nos résultats qu ‘une proportion importante des
personnes infectées d’une IST n’est pas traitée à temps ni correctement. Plusieurs raisons
sont à l’origine : 1) les personnes infectées, surtout les femmes, cherchent de l’aide médicale
assez tard et pas dans des structures médicales les plus efficaces ; et 2) l’algorithme utilisé
pour le diagnostic de l ‘écoulement vaginal n’est pas très efficace. Des stratégies
innovatrices pour résoudre ces questions sont nécessaires.
Résumé 133

Nous avons démontré que le coût du dépistage de syphilis est élevé et en même temps n’est
pas très efficace pour détecter tous les cas. Le traitement présumé de toutes les femmes
enceintes ainsi que leurs partenaires pourrait être considéré. Le traitement de masse des
travailleuses du sexe semble faisable.
Résumé 134

S AMENVATTING
Seksueel Overdraagbare Aandoeningen (SOA) zijn waarschijnlijk altijd al aanwezig geweest.
Het recente uitbreken van HIV/AIDS heeft desondanks een nieuwe dimensie aan SOA
gegeven en heeft bovendien de enorme culturele last benadrukt van ziekten geassocieerd
met seksueel contact.
Dit werk richt zich op de epidemiologie van SOA bij verschillende bevolkingsgroepen in
Nairobi, Kenia. We volgden vrouwen in prenatale klinieken en materniteiten, mannen en
vrouwen die behandeling zochten voor SOA klachten, mannen en vrouwen die behandeling
zochten voor algemene klachten alsook vrouwelijke “seks werkers”.
Een transversale studie van 520 vrouwelijke patiënten in de SOA referentie kliniek vond
plaats om de SOA etiologie te bepalen alsook de aanwezigheid van cervicale intraepitheliale
neoplasie (CIN). De prevalenties waren: HIV 29%, candidiase 35%, trichomoniase 25%,
bacteriële vaginose 16%, gonorroe 6%, chlamydia 4%, syfilis 6%, genitale wratten 6% en
genitale ulcera 12%. CIN werd aangetroffen in 12%. Factoren in verband met seksueel
gedrag, vooral het aantal seks partners, was geassocieerd met verscheidene SOAs.
Gonorroe, bacteriële vaginose, CIN en genitale wratten/ulcera waren geassocieerd met HIV
infectie. Deze gegevens duiden op een alarmerende prevalentie van SOA/HIV, vooral in de
jongere leeftijdsgroepen, op een gebrek aan “safe sex” methoden, en nood aan agressievere
preventie campagnes. Risico factoren voor SOA waren vooral het aantal seksuele partners.
Seksueel gedrag en “health-seeking” gedrag zijn belangrijke componenten van SOA
controle. Om dit gedrag te bepalen deden we een studie van 471 patiënten in de SOA
referentie kliniek, en een soortgelijke studie van 555 patiënten in eerstelijns klinieken in
lagere socio-economische wijken. We vonden dat een groot percentage van deze patiënten
al behandeling had gezocht in zowel de publieke als private sector alvorens naar de SOA
kliniek te komen. Er waren grote verschillen tussen mannen en vrouwen in het uitstellen van
medische behandeling. Daarenboven hadden vrouwen vaker seks terwijl ze symptomatisch
waren, vooral met hun vaste partner. Condooms werden zelden gebruikt tijdens ziekte.
Mannen gaven toe buitenechtelijke relaties te hebben in 68% en te betalen voor seks in
30%, maar gaven toch hun vrouw de schuld voor hun SOA. In de eerstelijns klinieken
vermeldden mannen buitenechtelijke relaties in 17% tegenover 8% van de vrouwen.
Vrouwen hadden een slechte kennis over gezondheid in het algemeen en SOAs in het
bijzonder. Significant meer mannen dan vrouwen hadden ooit al een SOA gehad. We
Samenvatting 135

vonden we een hoge prevalentie van gonorroe (3%) en chlamydia (6%), zonder onderscheid
van geslacht.
We hebben de validiteit van verschillende algoritmes voor de diagnose van gonokokken
chlamydia infecties bij 621 zwangere en niet zwangere vrouwen met klachten van vaginaal
verlies bestudeerd in 3 klinieken. The gemiddelde leeftijd was 24 jaar en 334 (54%) was
zwanger. De prevalenties waren: 50% candidiase, 23% trichomoniase, 9% bacteriële
vaginose, 7% gonorroe, 9% chlamydia, 7% syfilis and 22% HIV. Bij niet zwangere vrouwen
was infectie met gonokokken en chlamydia significant geassocieerd met 1) demografische
en gedrag risico factoren zoals vrijgezel zijn, jonger dan 20 jaar, meerdere seksuele partners
in de voorbije 3 maand; 2) het symptoom koorts; en 3) tekens zoals aanwezigheid van gelig
of bloederig vaginaal verlies, cervicaal mucopus, cervicaal erytheem en friabiliteit. Bij
zwangere vrouwen waren alleen jeugdige leeftijd, dysurie en koorts significant geassocieerd
met cervicale infectie. Maar geen enkel van deze variabelen was op zichzelf genoeg sensitief
en specifiek voor de diagnose van cervicale infectie. Verscheidene algoritmes werden
gecreëerd en op de studie gegevens toegepast. Het algoritme met risico factoren presteerde
iets beter dan het actuele Keniaanse algoritme. We besloten dat SOA een belangrijk
probleem vormt in Nairobi en zodoende moet worden aangepakt. Geen enkel van de geteste
algoritmes voor de behandeling van vaginaal verlies zou een duidelijke verbetering zijn van
het bestaande algoritme. En dus betere detectie methodes voor de specifieke etiologie van
vaginaal verlies zijn dringend nodig.
Een evaluatie van de routine gegevens van het syfilis screening programma toon taan dat de
syfilis serologie bij zwangere vrouwen lijkt te zijn gedaald van 7.3% naar 3.8% tussen 1993
en 1997. We deden ook een analyse van de routine data in 1998 om de prestatie en de kost
van een gedecentraliseerd syfilis-screening programma bij zwangere vrouwen in Nairobi te
beschrijven. De syfilis seroprevalentie 3.4%. Het programma slaagde erin vrouwen en
partners te screenen en te behandelen, maar faalde in het correct uitvoeren van de RPR
test. De kost per voorkomen geval werd berekend op 95 tot 112 dollar. De houdbaarheid op
lange duur van dit werkintensief programma is bedreigd door de hoge kost en de logistieke
vereisten. Alternatieve strategieën, zoals massa behandeling van alle zwangere vrouwen
moet worden beschouwd.
Als onderdeel van een grotere studie i.v.m. neonatale gevolgen van syfilis tijdens de
zwangerschap, hebben we de determinanten van partner notificatie bij zwangere vrouwen
beschreven. Van de 12,414 vrouwen getest voor syfilis bij bevalling waren er 377 (3%) RPR
reactief. Slechts 53% van deze vrouwen hebben hun partner ingelicht en 36% daarvan
Samenvatting 136

werden behandeld. Een negatieve zwangerschapsafloop was geassocieerd met niet
behandelen van de partner tijdens de zwangerschap (2% versus 15%, OR 6.1, 95% CI 0.8-
45). Deze studie toont lagere waarden voor partner notificatie/behandeling bij de bevalling
dan tijdens de zwangerschap. Dit suggereert dat voorlichting die een focus heeft op de
gezondheid van het ongeboren kind, een positief effect heeft op partner notificatie.
Innovatieve en lokaal aangepaste strategieën voor partner notificatie zijn nodig in de strijd
tegen HIV/SOA.
Een placebo gecontroleerde trial van regelmatige antibiotica profylaxis met azithromycin ter
preventie van SOA en HIV in een cohorte van seks werkers werd gestart in 1998. We
beschreven de baseline karakteristieken van 318 vrouwen die gescreend werden voor
opname in de studie. De gemiddelde leeftijd was 32 jaar en de gemiddelde duur van
sekswerk 7 jaar. Het gemiddeld aantal klanten per dag was 4. Hoog risico gedrag was
frequent: 9% had anaal seksueel contact, 19% seks tijdens menstruatie, en 3% gebruikte IV
drugs. Hoewel 20% vermeldde condooms te gebruiken met alle klanten, zei 37% nooit
condooms te gebruiken. Ondanks dat hoog risico gedrag was de prevalentie van SOA en
HIV relatief gezien laag: HIV 27%, bacteriële vaginose 46%, trichomoniase 13%, gonorroe
8%, chlamydia 7% en syfilis 6%. CIN was niet frequent (3%) en neigde ernaar meer
aanwezig te zijn bij HIV geïnfecteerde vrouwen. Vaginaal douchen was zeer frequent in deze
populatie en was significant geassocieerd met bacteriële vaginose. In deze studie hebben
we een grote groep HIV negatieve vrouwen met een hoog risico gedrag geïdentificeerd. Het
lijkt mogelijk deze vrouwen te benaderen met preventie programma’s, inclusief een trial van
HIV preventie via SOA chemoprofylaxis.
Een algemeen besluit van deze verschillende studies is dat SOA in Nairobi al verspreid zijn
van de hoge risico groepen naar de algemene bevolking waarbij mannen een belangrijke rol
spelen als “bridging” groep tussen seks werkers en hun vaste partners. Interventies speciaal
gericht op mannen om hen een meer verantwoordelijke houding ten opzichte van seksueel
gedrag aan te leren, terwijl ook interventies in de hoge risico groepen behouden blijven, lijkt
noodzakelijk.
We kunnen ook besluiten dat een groot aantal individuen met een SOA niet op tijd noch
correct behandeld worden. Er zijn verschillende oorzaken hiervan: 1) medische verzorging
voor SOA wordt lang uitgesteld, vooral door vrouwen, en wordt gezocht bij niet getrainde
gezondheidswerkers; en 2) het algoritme gebruikt voor de diagnose van vaginaal verlies
geeft geen goede resultaten. Innovatieve strategieën om deze problemen aan te pakken zijn
dringend nodig.
Samenvatting 137

We hebben aangetoond dat de kost van het screeningsprogramma voor syfilis in Nairobi
hoog is en tezelfdertijd niet efficiënt in het detecteren van de positieve gevallen. Empirische
behandeling van alle zwangere vrouwen en hun partners moet worden overwogen. Massa
behandeling van seks werkers is haalbaar.
Samenvatting 138

A CKNOWLEDGEMENTS
I am indebted to Prof Dr Marleen Temmerman who gave me the opportunity to complete this
PhD thesis. She stimulated me first of all to take on the job in Kenya, then to organize, set up
and implement research activities, and finally to publish the results. Without her enthusiasm
and guidance for scientific research in general and for reproductive health in particular, I
would not have succeeded.
Prof Temmerman’s team of the International Center for Reproductive Health (ICRH) at the
Department of Obstetrics and Gynecology, Ghent University, supported me all the way
trough. They have been fine colleagues and friends.
I am grateful for having had the opportunity to work with numerous world-known researchers
from the University of Nairobi, the University of Manitoba, the University of Washington, the
University of Oxford and others. It has been an incredible privilege. And I am indebted to a
long list of colleagues and staff in Kenya, many of who have become great friends, for their
continuous support.
I wish to recognize the support of the European Commission, who funded most of the work. I
thank Dr Lieve Fransen for her personal commitment to this work.
I thank the members of the reading committee for the helpful comments: Prof Dr M Dhont
(UG), Prof Dr J Plum (UG), Prof Dr A Meheus (UIA) and Dr Philippe Mayaud (LSHTM).
Special thanks go to my husband Nico and my family without whose support I would not
have had the courage and perseverance to bring this thesis to a good end.
And finally, I wish to dedicate this work to the people I have worked with and who have since
died of AIDS. Their dedication to fight the disease during their lifetime will always be an
example to me.
Thank you all.
Acknowledgements 139