fll l360 THE SCI(NCE OF IMPlOVIJIIC uvU 
- ----r----------
ii
Acknowledgements 
The authors would like to thank Kenyatta University (KU), the project leader, Prof. Michael Wainaina,
the project assistants Linda Amenya and Joseph Wahome; Mercy Gitau, Lisa Albert, Rick Homan,
Maureen Kuyoh, Susan McIntyre, Marsden Solomon and Jennifer Liku from FHI 360 for their
contributions to the design, analysis, assistance with the assessment, and review of the final report;
and Ruth Gathu for formatting the report designing the cover page; the moderators, volunteer trainers
and representatives of the PCEA Kariobangi South, PCEA Lang’ata, PCEA Nairobi West, Kingdom Life
Centre (KLC) Ngara, Adventist Centre for Care and Support (ACCS), NPC Parklands and PCEA Bahati for
their support in sensitizing the community on the project, conducting the trainings and mentoring the
house girls.  Special thanks go to the research assistants and analysts who worked tirelessly to collect
and document the findings of this study. We would also like to thank the house girls who opened their
lives to us by agreeing to participate in the HELP project and in the follow up research and the
employers for allowing the girls to participate and making alternative arrangements to cover for the
girls while they attended the training.  
Finally, we would like to thank the church members who besides being employers of the house girls
gave the project time during their church service to provide information on the project and at the end
of the training supported the project in organizing for graduation ceremonies for the house girls.  
This assessment was funded by the United States Agency for International Development’s (USAID)
Presidential Emergency Fund for AIDS Relief (PEPFAR), under two different cooperative agreements,
GPO‐A‐00‐05‐00022‐00 (the CRTU) and GHO‐A‐00‐09‐00016‐00 (the PTA). The information contained
in the publication does not necessarily reflect KU, FHI 360, or USAID policies.
  
iii
Executive Summary 
With funding from the United States Agency for International Development’s Presidential Emergency
Fund for AIDS Relief (USAID/PEPFAR), FHI 360 (formerly known as Family Health International) and
Kenyatta University (KU) carried out an evaluation of the effect of the house girls Health and Life skills
project (HELP) on the house girls’ vulnerability towards STI/ HIV and unintended pregnancies between
July 2010 and August 2010.  
In‐depth interviews with 18 church members and 27 employers of house girls, and face‐to‐face survey
interviews with 166 house girls aged 16‐24 from 7 project church sites in Nairobi, revealed that house
girls shared the same demographic factors as other house girls that have been demonstrated to
increase vulnerability towards HIV/AIDS and unintended pregnancies. The house girls in the project
had poor socioeconomic background and largely isolated with Sunday being the only day they were
allowed to leave the employers house and attend church service. However, compared to their peers
(from previous studies with house girls), the house girls who participated in the project had more
appreciation of their work, were confident, had improved on their communication skills and had
increased or initiated saving.  
At the end of the project, house girls demonstrated increased knowledge of modes of transmission and
prevention of HIV/AIDS, reproductive health indicators and indicated taking precaution to prevent
unintended pregnancy at last sex with majority indicating they used condoms. However, knowledge on
the woman’s menstrual cycle remained low.  
On the training, the modules on STIs and self‐esteem were indicated as the most beneficial to the
participants who attributed the improved communication skills, appreciation of their work and
increased confidence to the training. Employers and church members also acknowledged that these
were positive outcomes of the training. As a result of the training the house girls indicated that they
were now saving more money and more than half indicated that they had since opened bank accounts
in their names. Employers’ concerns were related to the time the girls spent at the training since they
had to get alternative help or do the work themselves. Additionally, the employers worried that as a
result of the training the girls may become more marketable and leave their employment.  
Overall, house girls, employers and church members would recommend the training to other house
girls, adding that the target group should be widened to include house girls above 24 years. Proposals
were also made to include topics on housekeeping, nutrition, cooking, baking, baby care, hygiene and
reading skills in the training.  
iv
Table of Contents 
ACKNOWLEDGEMENTS .......................................................................................................................II
EXECUTIVE SUMMARY .......................................................................................................................III
TABLE OF CONTENTS..........................................................................................................................IV
ACRONYMS .........................................................................................................................................V
INTRODUCTION...................................................................................................................................1
INTERVENTION ............................................................................................................................................. 1
RATIONALE.................................................................................................................................................. 2
GOALS AND OBJECTIVES......................................................................................................................3
STUDY DESIGN ....................................................................................................................................3
STUDY SETTING ............................................................................................................................................ 3
STUDY POPULATION ...................................................................................................................................... 4
STUDY METHODS .......................................................................................................................................... 5
House girl identification and screening ............................................................................................... 5
Employer identification/ selection....................................................................................................... 5
Church members’ identification/ selection.......................................................................................... 6
INTERVIEW SELECTION AND TRAINING ............................................................................................................... 6
DATA MANAGEMENT AND ANALYSIS ................................................................................................................. 6
VI. RESULTS....................................................................................................................................7
V. DISCUSSION ............................................................................................................................... 11
VI. CONCLUSION .......................................................................................................................... 13
REFERENCES...................................................................................................................................... 14
APPENDIX A: TABLES REFERENCED IN REPORT .................................................................................. 15
APPENDIX B: RESULTS OF THE FORMATIVE ASSESSMENT.................................................................. 27
APPENDIX C: HOUSE GIRLS PHASE 2 BASELINE SURVEY, SUMMARY OF FINDINGS............................. 30
  
v
Acronyms 
ACCS   Adventist Center for Care and Support
FBO   Faith Based Organizations
FHI 360 Family Health International
HELP   Health and Life skills Project
HIV   Human Immuno‐deficiency Virus
IUCD   Inter Uterine Contraceptive Device
KES   Kenyan Shillings
KLC   Kingdom Life Centre
KNH‐ERC Kenyatta National Hospital Ethics and Research Committee
KU   Kenyatta University
NPC   Nairobi Pentecostal Church
PHSC   Protection of Human Subjects Committee
PCEA   Presbyterian Church of East Africa
PEPFAR Presidential Emergency Plan for AIDS Relief
RA   Research Assistant
RH     Reproductive Health
STI   Sexually Transmitted Infection
SRH   Sexual and Reproductive Health
TA   Technical Assistance
USAID   United States Agency for International Development
VCT   Voluntary Counseling and Testing
WEL   Women’s Empowerment Link
WHO   World Health Organization
1
 Introduction 
In 2005, FHI 360 (then known as Family Health International) and Kenyatta University (KU) conducted a
qualitative formative assessment on the vulnerability of house girls to HIV infection and unintended
pregnancies. The study was conducted with 18 house girls aged between 18 and 22 years at
Presbyterian Church of East Africa (PCEA) Bahati Martyrs Parish. House girls were identified as a
population particularly vulnerable to HIV infection and unintended pregnancy because of: 1) their
socioeconomic background; 2) isolation and lack of social support; 3) the low status of their work; and,
4) their previous experiences with sexual coercion and violence. The formative research revealed that
house girls have some knowledge of modes of transmission and prevention of HIV; however,
knowledge and use of contraception including condoms were low.1
To validate the formative research findings and further inform intervention design, a baseline survey
was conducted in 2007 with 153 house girls 16 to 22 years of age. The survey corroborated formative
research findings on house girls’ knowledge of HIV, STIs, and unintended pregnancies. However, the
survey findings also showed that except for their isolation and hence limited access to health
information and services, house girls’ level of risk of STIs including HIV and unintended pregnancies
was similar to risk reported in other studies for other girls in this same age group2
. Consequently, an
intervention was designed in collaboration with Kenyatta University and PCEA Bahati Martyrs Parish to
offer house girls health and life skills training to reduce their vulnerability towards STIs including HIV
and unintended pregnancies.
Intervention
The intervention, referred to as house helps Health and Life skills project (HELP), was implemented in
September 2009 to June 2010 in collaboration with interested faith‐based organizations (FBOs) and
involved raising awareness among congregants of the FBO and the surrounding community, as well as
offering health and life skills training on STIs including HIV and unintended pregnancy prevention to
house girls. It was designed to be accessible and friendly to house girls and the demands of their job.
Therefore, training was free for all participating house girls and is conducted on Sunday afternoons
since many house girls get time off on Sunday. The training sessions were held at host FBO sites for 2½
hours on Sunday afternoons for a total of 24 Sundays. Additionally, 1‐hour interactive radio talk‐shows
are held on local FM stations, with opportunity for listeners to call in or send text messages during the
show. The girls were taken through a seven module curriculum covering 1) Self‐esteem and
communication skills; 2) Sexual and reproductive Health (SRH); 3) STIs including HIV prevention and
protection; 4) Sexual violence, alcohol and drug abuse; 5) Personal savings and financial management;
6) Basic household safety and security; and, 7) Basic first aid.  
HELP was first piloted at PCEA Bahati Martyrs Church in Nairobi between December 2007 and August
2008 with funding support from President’s Emergency Plan for AIDS Relief (PEPFAR). Of the 50 girls
registered, 44 completed training and were certified. House girls who participated in the training
reported increased knowledge on how to protect themselves and how to make responsible decisions
to avoid STIs including HIV and unintended pregnancies; and increased competence and improved
                                                      
1
 More details on the results of the assessment are available in Appendix B.
2
 A summary of the baseline survey results is available in Appendix C.
2
relationships at work. Employers have also reported seeing positive changes in their house girls
particularly in their communication skills and responsibility in the household. The 44 house girls who
successfully completed the training have since been trained as peer educators and are currently
reaching out to their peers with STIs including HIV and unintended pregnancy prevention messages.  
In 2008 and 2009, the project received additional funding to generate more lessons and expand the
intervention to six other sites in Nairobi namely; PCEA Kariobangi South, PCEA Lang’ata, Kingdom Life
Centre (KLC) Ngara, Adventist Center for Care and Support (ACCS), Nairobi Pentecostal Church (NPC)
Parklands, and PCEA Nairobi West. In total, the project was undertaken in seven sites including PCEA
Bahati3
. Training of the house girls began in September 2009 and ended in June 2010. As part of the
training, pre‐ and post‐ training tests were conducted to assess knowledge on each of the seven topics
taught.  
The seven sites enrolled in the project were divided into technical assistance sites (TA sites) and
replication sites. In TA sites, volunteer trainers are identified from the congregation or sites and are
invited for a five day training of trainers’ course where they are taught facilitation skills and organized
into training teams based on their strengths on the 5 modules that the house girls are taught. The
teams then go back and prepare their lesson plans and conduct trainings with the house girls. The
volunteer trainers receive technical support from the pool of expert trainers attached to the
intervention on a regular basis and all training materials save for the training venue are provided by
the program. At the replication sites, the trainings are conducted using experienced trainers attached
to the project with the assistance of volunteer trainers recruited from the church and the project
supports all other activities as in the TA sites save for venue. In comparison, providing technical
assistance is cheaper in the long run and it was anticipated it would enable the churches to take over
full project implementation while replication is more expensive and would be used in sites with
minimal technical capacity.
Rationale
There is little research that has been conducted on the situation of house girls and through this
research the experiences and learning from this unique intervention can be documented and used to
inform future programming. The intervention was implemented with technical support from KU and
FHI 360. It was planned that in the future the churches would be able to take on full implementation of
the intervention, hence the introduction of the two approaches.  
The intervention was intended to improve house girls’ knowledge and utilization of STIs including HIV
and unintended pregnancies prevention approaches. It was to enable the house girls to identify risks
and take appropriate actions to reduce their vulnerability of STIs including HIV infection and
unintended pregnancies. The further scale up of this intervention will depend not only upon its
effectiveness to promote usage of STI including HIV and unintended pregnancy prevention methods,
but also upon its acceptability among stakeholders. Therefore, the study will incorporate in‐depth
interviews with church leaders/ members and employers of house girls. The church leaders are
essential in obtaining buy‐in from their congregants and supporting the intervention through provision
of training hall and facilitating church announcements while the employers directly influence the
decision of the house girls to attend the interventions held every Sunday and would ideally note
                                                      
3
 PCEA Bahati is continuing programmatic activity with technical support from the project team
3
changes in the house girls attitude and behavior particularly as relates to basic household security and
first aid.  
Goals and Objectives 
The overall goal of this evaluation was to describe the effect of the intervention on house girls’
knowledge and practices as relates to HIV and unintended pregnancy prevention.
Specifically, the objectives were:
1. To examine house girls’ sources of vulnerability to STIs including HIV and unintended
pregnancy, their post training knowledge of these subjects, and their sexual and reproductive
health behaviors.
2. To describe house girls’ capacity to take risk‐reduction action on STIs including HIV and
unintended pregnancies
3. To descriptively compare the outcomes of the intervention in areas characterized by
differences in the socioeconomic status of the population
4. To determine perceptions and attitudes towards the program by the house girls, their
employers and church members.
Study Design 
This evaluation used a combination of quantitative and qualitative methods. A post‐ intervention, cross
sectional survey was conducted with house girls aged 16 – 24 years who participated in the
intervention. In addition, in‐depth interviews with employers of house girls who participated and
church members at the sites where the trainings took place were also conducted. The evaluation was
approved by FHI360’s Protection of Human Subjects Committee (PHSC) and the Kenyatta National
Hospital Ethics and Research Committee (KNH‐ERC).
Study setting
The intervention was implemented in five out of eight constituencies in Nairobi (Figure 1). The study
was conducted at all seven project sites located in the five constituencies including PCEA Kariobangi
South, PCEA Lang’ata, PCEA Nairobi West, Kingdom Life Centre (KLC) Ngara, Adventist Centre for Care
and Support (ACCS), NPC Parklands and PCEA Bahati. Together the seven sites reach between 16,000
and 17,000 congregants; PCEA Bahati has the largest congregation and PCEA Lang’ata and KLC Ngara
the least as indicated in Table 1.The training sessions took place from November2009 to July2010 and
the evaluation followed in August2010.
  
4
Table 1:  Estimated congregation in the targeted churches  
Name of church Constituency Estimated congregation4 Number of targeted girls
PCEA Lang’ata Lang’ata 500 50
KLC Ngara Starehe 500 50
NPC Parklands   Westlands 1,000 50
ACCS   Westlands 4,500 50
Kariobangi South   Embakasi 1,800 50
PCEA Nairobi West   Lang’ata 700 – 800   50
PCEA Bahati Makadara 7,000 – 8,000 50
Total     16,000 – 17,100 350
Figure 1: Map of project site location  
Study population
The three target groups for this study were house girls, employers of house girls and church members
who volunteered to assist with the training sessions. For a house girl to be eligible to participate in the
study she needed to meet the following criteria:
                                                      
4
 Estimates provided by the Reverend/ Pastor in charge of the churches
5
1. Completed the intervention training at one of the intervention sites
2. Aged 16 – 245 years
3. Employed as a house girl in Nairobi;
4. Living with employer;
5. Not planning on leaving employment in the next year;
6. Affiliated to the church either directly through their employer or themselves being church
members or indirectly through referrals by other church members not their employer.
All employers whose house girls participated in the intervention and whose house girls consented to
them being interviewed were eligible for inclusion in the in‐depth interviews. The church members
who were volunteer trainers or other church leaders were also eligible for inclusion in the in‐depth
interviews.
Study methods
House girl identification and screening
The house girls to be included in the survey participated in the house helps’ project and consequently
attended training sessions as part of the project. The survey was introduced to the girls on the last day
of their training where they were given information about the objectives of the survey, its implications
and their role in the survey before being invited to participate.  
The Research Assistants (RA) subsequently made appointments with those willing to participate to be
interviewed at a specific pre‐arranged time that was convenient for the house girl. The participants
who respected their appointments were reimbursed at a standard flat rate of 200 Kenya shillings
(estimate US$2.706
) for their travel. Oral Informed consent was administered to all eligible participants
who presented themselves at the interview locations and those who consented were interviewed. A
structured questionnaire was used to collect information and covered the seven topics from the
training sessions. Questions assessed knowledge from the sessions, related behaviors, social support
systems, self‐efficacy and perceptions of the training sessions.  
The intervention with house girls targeted 50 house girls per site bringing the total number of house
girls expected to enroll to participate in the project to 350. It was assumed that the project would have
a completion rate of 85% and that 80% of the girls approached would accept to participate in the
study, thus a total number of 238 girls were expected to participate in the study. However, the project
was affected by the Christmas break which occurred midway through the training program. The break
resulted in a large dropout of house girls from the project as they took up employment with new
employers. Hence the total number of house girls who participated in the study was 166.
Employer identification/ selection
Once the house girls were interviewed they were asked whether it would be okay to interview their
employers. The participants who consented to have their employers interviewed were given a consent
package composed of a brief on the survey and its objectives, a request to conduct an interview and a
                                                      
5 Although the baseline protocol had anticipated reaching house girls 14‐22 years of age, the actual survey reached a minimum age of 16.
In all intervention sites so far we have not come across house girls below this age. The upper age limit is set at 24 to include those who
had expressed interest to join an expanded phase of the study, following media publicity of the pilot phase certificate award ceremony at
PCEA Bahati in August 2008.  
6 Exchange rate 1 USD = KES 73.94 (as at 7th April 2010)
6
form for telephone contacts for those willing to be contacted for interviews to take home to their
employers.  
The house girls were asked to bring back the duly signed forms which were then used to randomly
select the employers to interview. A transport reimbursement of Kenya shillings 200 was given to the
house girls who returned the duly signed forms with employer contacts. A total of 27 employers were
randomly selected representing three to four per site. A semi‐structured questionnaire was used to
assess employer perceptions of the training, what they liked and did not like about it and their
thoughts on the effect it had on their house girl.  
Church members’ identification/ selection
Church leaders, church project coordinators, and church volunteer trainers were purposively selected
for interviews depending on their availability and willingness to be interviewed. A total of 18 church
members were randomly selected, i.e. one or two volunteer trainers and one church
leader/coordinator per site.
Interview selection and training
The survey employed 10 research assistants (RAs) for the duration of data collection. The RAs had basic
counseling skills and were required to complete the FHI360 research ethics curriculum training which
was further reinforced during a five day survey specific training. The RA training also covered survey
instrument and informed consent administration, participant identification and general interviewing
techniques. RAs were trained to be particularly aware of the nuances of conducting surveys with
youth, and to ensure that there is full compliance with concepts of privacy and confidentiality. A half‐
day of training was devoted to sensitization in such nuances, and conducted by the Women’s
Empowerment Link (WEL).7 The same organization also provided on‐site counseling support in case
survey respondents get so affected by interview questions as to need such support. RAs were
sensitized during training to be able to identify such needs as they arise during individual interview
sessions. The same RAs also conducted the in‐depth interviews and had special training on handling
open‐ended questions, probing and prompting.  
Data management and analysis
Survey data were entered into an Epi‐Info database and converted to SPSS version 17.0 for analysis.
Descriptive analysis was conducted and presented in tables showing the distribution of responses to
each question. No statistical tests were performed. Open‐ended questions were categorized and the
most frequent responses presented.
In‐depth interviews were recorded on paper using pens as well as audio‐taped. The discussions were
then simultaneously transcribed and translated into English. The employer data was then transferred
into SPSS while the church member data was transferred to Excel. The responses were summarized
and the most frequent responses identified. Divergent responses are also reported to present a full
range of opinions.
  
                                                      
7
WEL is a non‐governmental, non‐profit making organization whose major role is to promote and uphold women and girls in society
through advocating for opportunities that explore the potentials of women and girls.
7
VI.RESULTS 
First we looked at some of the characteristics of the house girl sample including their demographic
information, religious and other activities that they undertake and their employment. Next, we
examined the sources of reproductive health information that they have access to and their general
reproductive health knowledge including knowledge of fertility, pregnancy prevention and STIs/HIV.
Third, we explored the reproductive health related behaviors of the house girls including their sexual
activity and experiences with sexual violence, contraceptive use and use of RH health services. The
next section explored their feelings about their social support system and their self‐efficacy. The final
section presented results on the training itself through the both the house girls and the employers
views on the training.
Socio‐demographics
The mean age of the house girls who were interviewed was 21.4 years with a range of 16 to 25 years
(Table 1).A little more than half had primary school education or less. Only 18% had completed
secondary school or gone on to post‐secondary education. Nearly all the house girls had never been
married though one‐fifth had ever had a child. Most house girls would like to have a child in the future
with most wanting to wait until after they are married before having one.
Religious and other activities
To effectively provide information and education to house girls, it’s important to know where they
spend their time outside of work. Religious venues appear to be a good way to reach large numbers of
house girls. The majority of house girls reported that they were Protestant, and most of the rest were
Catholic (Table 2).Nearly all of them attend church services with most attending once per week. Youth
clubs may be another possible way to reach house girls since nearly one‐fourth said that they were
members of such a club.  
Employment
On average, the house girls in the survey had been with their current employer for 2.8 years (with a
range of 1 to 5 years) though had worked as a house girl for an average of 3.4 years (range of 1 to 5)
(Table 3).Close to two‐thirds of the house girls get one day off per week and many others also get
holidays or two or more days off per week. About five percent report that they do not get any days off
or only get holidays. The average income per month in their current position is KES 3,221 (US$43.56).
Most girls send some money home to their family and on average they send about half of their income.
About 90% save money; about half have a bank account in their own name.
Sources of information  
Media sources are good ways to reach large segments of the population with information on
reproductive health. Most house girls had read a newspaper and listened to the radio at least
occasionally; 14.5% and 7.8% respectively reported that they never read a newspaper or listen to the
radio (Table 4).The biggest source of reproductive health information for these house girls was the
course at the church (64%). But aside from the church, nearly half received this information from the
radio and from television. Far fewer report receiving RH information from a newspaper.
8
Reproductive health knowledge
Reproductive health education was a major component of the HELP training workshop. On the survey,
the house girls were asked five basic questions to assess their RH knowledge. Knowledge of some
questions was good though gaps were evident (Table 5).Ninety six percent could name at least one
modern method of contraception (over 90% could name two and half knew four methods).Most (90%)
knew that a girl could get pregnant the first time she has sex. Over three‐fourths (82%) knew that a girl
could get pregnant if she has sex standing up. Little more than two‐thirds correctly stated that
pregnancy was still possible even with withdrawal while few (29%) correctly identified the point in a
woman’s menstrual cycle when she was most likely to become pregnant. For those who named specific
modern methods of contraception (pills, IUCD, injectables, male condom, male sterilization and female
sterilization) they were asked to describe how to use the method; more than 75% could correctly
describe the pill, injectables, male condoms and male sterilization. Fewer could explain how the IUCD
works (64%) and very few could explain female sterilization (20%). All of the house girls got at least one
response correct but only 13% got all five. Most got three or four correct (data not shown).  
Knowledge of STIs including HIV
Some gaps were also evident from the results of their STI knowledge. Some of the main symptoms of
STIs include abnormal discharge, itching or burning in the pelvic region, sores or warts on the penis,
painful urination and pain during sex. Table 6 shows that while just over half named itching and sores
as symptoms fewer could name some of the other symptoms. All the house girls interviewed had
heard of HIV and nearly all know it can be acquired through sexual intercourse. They were less likely to
state some of the other ways that HIV can be acquired such as through sharing of needles and medical
equipment; blood transfusions and from mother to child through birth or through breastfeeding. Most
house girls cited abstinence and condom use as the main ways to avoid HIV. These were also the two
main ways cited to avoid all STIs (data not shown). Most (89%) know that condoms offer protection
from both HIV and pregnancy.
Sexual activity and sexual violence  
Sexual activity was fairly widespread among this group though most did not appear to be having
multiple partners. Nearly two‐thirds of the house girls interviewed had ever had sexual intercourse. A
similar percent reported that they had a boyfriend at the time of the interview; of those with a
boyfriend, 56% reported they had had sex with him. Of those who have had sex, the average number
of lifetime sexual partners is 1.7 with an average of 0.97 partners in the past year.
Among those who had had sex, 8% reported that they had received a reward or gift in exchange for
sex. About one‐fourth of those who had had sex said that they had been forced to have sex
(representing 15% of the total study population) and of these house girls, 15% (4 house‐girls) said that
an instance of forced sex had happened in their employer’s home (not necessarily in their current
employer’s household).  
Contraceptive use
Contraceptive use was high among the sexually active house girls with 79% reporting that they had
done something to avoid pregnancy and 74% reporting they had done something the last time they
had sex. The most common means used to prevent pregnancy at last sex was male condoms though
pills and injectables were also used by about 4% (though 10% have ever used each of these methods).
Not all the house girls are relying on modern methods and 10% reported that they avoided pregnancy
9
at last sex by relying on safe days or periodic abstinence. For those who didn’t use a contraceptive
method the last time they had sex the main reasons why they did not were because: they don’t know
what they are; they didn’t think about it; their partner does not like methods or they wanted to get
pregnant (data not shown).  
RH and other services sought  
Over half of the study sample had sought out selected RH services in the past 12 months with VCT
being the service most commonly used (Table 9). Only a few sought out services for family planning,
rape or violence. Almost all of the house girls knew where to go if they want to get tested for HIV with
the greatest numbers saying they would either go to a hospital or the VCT new start center. Nearly
three‐fourths have already been tested for HIV. Only 12 house girls reported that they had had an STI
symptom in the last 12 months; of these eight girls went to the hospital and four said they did nothing
and it went away without treatment (data not shown).
Social support and self‐efficacy
The house girls were asked 14 questions to assess their feelings about their social support network
(Table 10).The results show that overall most of the girls in the survey feel that they have a strong
social support system. Between 60‐ 88% either agreed or strongly agreed with statements that reflect
being cared about or respected by people in their lives e.g. their employer, employer’s spouse, children
and relatives; their church and their community. The statements where the fewest house girls agreed
or strongly agreed were about their employer’s spouse or employer’s relatives caring about them, but
higher percentages felt that these two groups respected them. For the house girls who had negative
responses about their social support system there was no evident pattern of certain girls having
virtually no social support; of the 14 questions only 5% had negative responses to at least half the
questions (and no one had more than eight negative responses)..  
In addition, house girls were asked eight questions to assess their self‐efficacy, or their confidence in
their ability to control who to have sex with and when, their use of condoms and their ability to seek
out health care or other services or information (Table 11).Between 83‐ 99% expressed that they were
either confident or very confident in their ability to self‐decide. The scenarios where house girls
showed the least confidence were in their ability to choose who to have sex with (16% not confident or
not at all confident) and to visit a reproductive health clinic (7%).Similar to the social support findings,
there was no pattern of certain girls having low self‐efficacy, in fact, 62% of the house girls showed
positive self‐efficacy on all eight questions and only two house girls had negative responses to more
than two questions (for both girls they showed negative self‐efficacy on four questions).  
HELP Training
The list of training topics the house girls attended is shown in Table 12.Participation was greatest in
“self‐esteem and communications” and “STIs including HIV prevention.” The topic of “Sexual violence,
alcohol and drug abuse” was attended by the fewest of those surveyed. Self‐esteem and STIs were the
two topics that house girls said that they benefitted from the most. Others that over 40% said that
they benefitted from include: personal savings and financial management; basic first aid, sexual and
reproductive health and basic household security. The majority did not name any topic as the one from
which they benefitted the least.  
10
For each individual topic, the majority of house girls reported that they are doing something different
as a result of having attending that particular session (Table 13). The most changes were reported as a
result of the sessions on self‐esteem and personal finance while the fewest reported changes were
from basic first aid and sexual and reproductive health. Nonetheless, at least half of the house girls
reported making some change as a result of every training topic. The main reported changes for each
topic are listed in Table 14.
The house girls’ employers who participated in the in‐depth interviews primarily learned about the
training through a church announcement (52%) or word of mouth (30%). The majority of employers
agreed to allow their house girl to attend the training because they thought it would benefit her,
providing her with education and skills. Most agreed that they thought the training was beneficial and
some specifically said it was “motivational”, “enlightening”, “empowering” or “informative”. Some of
the ways they saw the benefits was in that “self‐ esteem went up”, “the house girl was grateful” and
“the house girl is more confident and enjoys her work. ”Some specific advantages that were noted by
more than one employer are better time management skills, increased knowledge about reproductive
health issues (e.g. HIV, STIs and pregnancy), she is saving more and her hygiene has improved. Several
also commented that their house girl now has an improved attitude about her work.
Of the 18 church members interviewed all played a role in the training program either as a volunteer
trainer, coordinator or liaison. All of the church members thought the training was a positive activity
and all would recommend it to others. One called the training an “eye opener” and another said that
“from the changes in the girls one can note how far they have come.” Yet another commented that it is
“good because very few groups think of house girls.” The main positive effects that the church
members noted about the training was improved communication skills including being more
empowered to talk, improved self‐esteem, a more positive attitude, improved knowledge and financial
management skills.  
No employers made negative comments about the training and few saw any disadvantages, though a
few did comment about the time it took away from the house or that the employer had to do the
cleaning herself on Sundays. Most, however, agreed that Sunday is a good day for the training though
there are some that would prefer the training in the morning and others who would prefer the
afternoon. All but one employer said they would recommend the training to other employers of house
girls.
While many church members also reported that there were no disadvantages to the training there
were others who said that some employers had indicated that they feared their house girls would
leave (escape or runaway from) their households after the training. Also, a few said that there were
comments from people that girls would pick up bad habits from each other or become disrespectful to
their employers.  
Almost all of the church members interviewed had recommendations for the training. Many said that
more time is needed to cover the topics. Several commented on increasing the age range and one
suggested that there should be different classes based on age; one each for 16‐18 year olds, 19‐24 year
olds and 25‐30 year olds. Many also recommended adding extra topics including housekeeping,
nutrition, cooking, baking, baby care, hygiene, and reading skills. Even though other topics were
requested, most said that they would keep all of the current topics in the training.
11
V. Discussion 
Although the study design was not pre‐  and post‐intervention design, the results suggest that the
training may have had a positive effect on the participants. The house girls reported good knowledge
and use of RH and contraceptives, strong social support and positive self‐esteem. Furthermore, the
house girls reported adopting positive behaviors such as initiating or increasing their savings and
opening bank accounts in their names.  
Sources of vulnerability
This study of house girls in Nairobi in the HELP project supported earlier findings that house girls’ had
poor educational and economic background, and were isolated which when combined with the low
status associated with domestic work increased their vulnerability to STI/HIV and unintended
pregnancies. Low social and economic status has also been associated with less ability to negotiate
condom use or to get out of abusive relationships (WHO, 2002). The house girls’ poor socioeconomic
status means that they might be more likely to fear a) losing one’s job, b) letting one’s family down, c)
destitution, and d) abandonment thus reducing their ability to negotiate for safe sex (Zierler & Krieger,
1997). However, it was established that this was not the case with house girls who had gone through
the HELP project. These house girls demonstrated higher self‐efficacy and confidence in their ability to
control who to have sex with and when.  
The house girls’ live with the employers and hardly had any free time to be with family and friends.
This parental absenteeism further predisposes them to HIV and unintended pregnancies. In a study
among urban youth who did not live with their parents showed that the youth were more likely to be
sexually active and have unintended pregnancies than their peers (Ngom, Magadi, & Owuor, 2003).
The minimum wage of a house girl in Nairobi as recommended by the government of Kenya at the time
of the study was 6,743/month (about $91.2/month) (Ministry of Labour, 2010). However, the project
established that most times the house girls received less than half the recommended amount (average
of $44/ month). Despite the low earnings, the house girls reported sending about half of their wages to
support their families (parents and siblings) back in the village which undoubtedly laid financial strain
on them. Nevertheless, high proportion of house girls reported saving money with half of them stating
that they had functional bank accounts in their names. The issues of finance remains important as
studies continue to demonstrate that money and other financial benefits impact sexual activity
(Nobelius, Kalina, Pool, Whitworth, Chesters, & Power, 2011). In introducing financial management as
one of the course for the house girls in the project, it was hoped that the girls would save or increase
their saving, gain financial independence and ultimately increase their ability to negotiate for safe sex.  
Ability to take risk reduction measures
Knowledge of reproductive health and STI/HIV was higher among the house girls compared to their
peers on most of the indicators including awareness of at least one modern method of contraceptive,
dispelling of myths on how pregnancy occurs, awareness of at least one mode of HIV transmission and
at least one sign and symptom of STIs. The proportion of those who were aware of the period during a
woman’s menstrual cycle when she is most likely to get pregnant was no different from that of their
peers (KDHS, 2008‐09; Buga, Amoko, & Ncayiyana, 1996). Another positive outcome of the project was
the higher contraceptive use among sexually active house girls. The study showed that a higher
12
proportion of house girls reported having had sex in the past (65%) compared to girls their age in the
general population (35%) (KDHS, 2008‐09). However, unlike the girls their age in the general
population (28%), the house girls reported higher incidence of taking precautionary measures to
prevent unintended pregnancy (74%)(KDHS, 2008‐09). As their peers, condoms and other short acting
contraceptives were the most used in pregnancy prevention. The study did not explore the consistency
of condom use among the house girls. That withstanding, condom use should be encouraged while
incorporating messages on dual method use and continuing education on the menstrual cycle to
enhance outcome of the barrier methods and short acting methods used. Additionally, shift from short
acting methods to Long Acting methods should be promoted amongst house girls who are sexually
active and have sex regularly.  
House girls reported the lowest self‐efficacy in their ability to access RH services. Thus, there is need to
explore ways of providing house girls with contraceptive services in a way they are comfortable
accessing. As a first step, the project could consider mapping out RH services in the region with a
special focus on facilities with Youth Friendly Services (YFS) and provide these as referral points during
training.  
Furthermore, about 15% of the house girls who have had sex had experienced an episode of sexual
coercion and violence, some of which occurred in the employers’ household (not necessarily current
employer). Past studies indicate that these girls are likely to experience subsequent incidents of forced
sex, as well as (consensual) sexual risk‐taking behaviors including multiple sexual partners and non‐use
of condoms, and consequently increased risk of unintended pregnancy and sexually transmitted
infections (STIs) including HIV   (Ganju, et al., 2004). Although this study establishes that there are
incidences of forced sex occurring in an employer’s household (not necessarily current employer), the
findings are not sufficient to conclude that the nature of the work puts the women at more risk of
sexual exploitation. However, what is clear is that the house girls are at risk of STI/HIV and unintended
pregnancies due to their past exposure to forced sex.  
Studies have shown that female adolescents with low self‐esteem tend to think of intimate
relationships as imperative to a positive self‐concept (Raiford, Seth, & Diclemente, 2012).
Consequently, interventions with adolescents have focused on increasing their self‐esteem and
strengthening their social support network. In this study, it was established that the house girls in the
project demonstrated self confidence in their ability to make decisions affecting their sexual activity
and had high self‐esteem.  
13
VI.Conclusion 
Overall, the study has demonstrated that the training provided an opportunity for the house girls to
receive information that otherwise proved elusive to them given the nature of their work. Improved
self‐esteem and efficacy and the increased financial management realized among the house girls have
been directly attributed to their participation in the project by the house girls themselves, their
employers and by the church members interviewed. Increased knowledge on STIs (including HIV
prevention) and prevention of unintended pregnancies was realized. After the training the house girls
said that their communication skills improved, they had better appreciation of their work and they
were more confident. These same observations were made by the employers and church members as
well. Reservations about the training as expressed by the employers included fear that the house girls
would leave once trained and the time spent by the house girls during training. All in all, employers,
church members and the house girls would recommend the project to others.  
14
References 
Buga, G. A., Amoko, D. H., & Ncayiyana, D. J. (1996). Adolescent sexual behaviour, knowledge and
attitudes to sexuality among school girls in Transkei, South Africa. East African Medical Journal 73 (2) ,
95‐100.
Ganju, D., Jejeebhoy, S., Nidadavolu, V., Santhya, K. G., Finger, W., Thapa, S., et al. (2004, June). The
adverse health and social outcomes of sexual coercion: Experiences of young women in developing
countries. Retrieved February 8, 2013, from Population Council web site:
http://www.popcouncil.org/pdfs/popsyn/PopulationSynthesis3.pdf
KDHS. (2008‐09). Kenya Demographic and Health Survey. Kenya: Kenya National Bureau of Statistics
(KNBS) and ICF Macro.
Ministry of Labour. (2010, May 1). Regulations of Wages Act. Kenya Subsidiary Legislation, Labour
institutions act, legal notice no. 97 . Nairobi, Kenya: Kenya Gazette.
Ngom, P., Magadi, M. A., & Owuor, T. (2003). Parental presence and adolescent reproductive health
among the Nairobi urban poor. Journal of Adolescent Health 33 , 369‐377.
Nobelius, A. M., Kalina, B., Pool, R., Whitworth, J., Chesters, J., & Power, R. (2011). Sexual partner types
and related sexual health risk among out‐of‐school adolescents in rural south‐west Uganda. AIDS Care
23 (2) , 252‐259.
Raiford, J. L., Seth, P., & Diclemente, R. J. (2012). What Girls Won't Do for Love: Human
Immunodeficiency Virus/Sexually Transmitted Infections Risk Among Young African‐American Women
Driven by a Relationship Imperative. Journal of Adolescent Health , E publication ahead of print.
WHO. (2002). Integrating gender into HIV/ AIDS programmes. A report of an expert consultation.
Geneva: WHO.
Zierler, N., & Krieger, S. (1997). Reframing women's risk: Social inequalities and HIV infection. Annual
Review of Public Health , 401‐436.
  
15
Appendix A: Tables referenced in report 
Table 1:Socio‐demographic characteristics of house girls
N Mean, Range or Percent
Age at last birthday (mean, range) 165 21.4 (16‐25)
Highest level of school (%)
Primary incomplete
Primary complete
Secondary incomplete
Secondary complete
Post‐secondary education
Vocational/Tertiary  
166  
8
50
22
16
1
2
Marital Status (%)
Never married
Divorced
166  
99
<1
Ever had a child (%) 165 20
Would like a/another child in the future (%) 164 94
When would like to have a child/ another child (%)
Now
After marriage
Other
2‐9 years
later/older
1 year
pregnant already
Refused/No response
147  
3
80
8
5
1
<1
<1
<1
Table 2:Religious and other activities in which house girls participate
N Percent
Religious Denomination  
Catholic
Protestant
Muslim
Other
165  
19
78
<1
3
How often church services attended:
Less than once a month
Once a month
Once a week
More than once a week
At least once a month
Don’t attend religious services
165  
<1
6
83
6
4
<1
Member of a youth club 165 23
  
16
Table 3: House girls’ employment
N Mean, Range or
Percent
Years of employment with current employer (mean,
range)
150 2.8 (1‐5)
Years employment as a house girl (mean, range) 152 3.4 (1‐5)
Days off (%)
None
Only holidays
One day per week
One day per week and holidays
Two days per week
Two to four days per week (with or without holidays)
Other
164  
4
2
63
13
14
2
2
Income per month (Kenyan shillings) (mean, range) 165 3,221 (1,000‐7,000)
Send home money to family (%) 166 95
Amount of money sent home to family last time sent
money home (Kenyan shillings) (mean , range)
157 1,550 (300‐5,000)
Save money through: (%)
Through groups (Mary go round)
Savings account
Employer
Do not save money
Other
Relative
Mpesa
Envelope/bag/piggy bank
Home bank  
Refused/ No response
166  
4
49
5
9
2
4
9
10
7
<1
Have a bank account in own name (%) 165 55
  
17
Table 4: Respondents’ access to and sources of RH information
N Percent
Frequency read newspaper
Every day/ Almost every day
At least once a week
At least once a month
Less than once a month
Never
Other
166  
26
41
11
6
14
2
Frequency listen to radio
Every day/ Almost every day
At least once a week
Less than once a month
Never
Other
166  
79
11
<1
8
1
Sources of information on sexual and reproductive health
Radio
Television
Newspaper
Magazine
Other house girls
Friends who are not house girls
Course at church  
Someone else at church
Employer
Health care provider  
Boyfriend
Family member
HELP course/training
None
165
166
166
166
166
166
165
164
166
166
165
165
165
162
47
46
17
4
8
14
64
12
10
11
2
6
10
1
  
18
Table 5: House girls’ reproductive health knowledge
N Percent
Correctly identified the point in a woman’s menstrual
cycle when she is most likely to become pregnant
164 29
Knows that a girl can get pregnant the first time she has
sex
165 90
Knows that a girl can get pregnant if she has sex standing
up
165 82
Knows that pregnancy is still possible even with
withdrawal
164 36
Knowledge of modern contraceptive methods:
Identified at least one modern way to avoid getting
pregnant*
Identified at least two modern ways to avoid getting
pregnant
Identified at least three modern ways to avoid getting
pregnant
Identified at least four modern ways to avoid getting
pregnant
165  
96
91
76
50
If named above, could explain how a person uses the
following methods:  
Pill
IUD
Injectables
Male condom
Male sterilization
Female sterilization
122
58
84
146
11
16
78
64
82
94
82
20
 *Modern methods are OC pills, IUCDs, injectables, diaphragms, male/female condoms,
implants, male/female sterilization, and emergency contraception.
  
19
Table 6: Knowledge of STIs including HIV
N Percent
Spontaneously identified signs and symptoms of an STI:  
Abnormal discharge
Itching or burning in pelvic region
Sores or warts on penis
Painful urination
Pain during sex
Don’t know/ remember  
166
165
166
166
166
154
48
56
60
38
16
8
Has heard of HIV 163 100
Believes HIV can be acquired through:  
Sexual intercourse
Sharing needles & medical equipment
Blood transfusions
From mother while in womb
Mother to child at birth
Mosquito or insect bite
From breastfeeding
Casual contact with infected person
Kissing (“deep kissing” or “kissing w/ open sore”)
Contact with open wound, blood, or other body
fluid
Don’t know
Other
166
166
166
166
166
165
165
165
166
166
149
166
99
77
58
4
24
1
23
19
20
8
3
5
Believes HIV can be avoided by:  
Abstinence
Staying faithful to one partner who was tested
Staying faithful to one partner
Encouraging partner to stay faithful
Avoiding contaminated blood
Using condoms
Avoiding sharing needles
Avoiding casual partners
Avoiding sex with sex workers
Avoiding anal sex
Nothing
Don’t know
Other
166
165
166
166
166
166
166
166
166
166
164
151
166
86
23
19
1
20
90
24
4
3
<1
0
<1
11
Knows condoms protects from HIV and pregnancy   166 89
  
20
Table 7: Sexual activity and sexual violence
N Mean, Range or
Percent
Ever had sexual intercourse (%) 166 65
Has a boyfriend (%) 166 64
If has a boyfriend, has had sex with boyfriend (%) 106 56
Number of sexual partners, among girls who have had sex
(mean, range)
105 1.7 (1‐5)
Number of sexual partners in the last year, among girls
who have had sex (mean, range)
103 0.97 (0‐4)
Has received reward or gift in exchange for sex, among
girls who have had sex (%)
108 8
Has been forced to have sex, among girls who have had
sex (%)
107 24
If forced to have sex, did any of these times happen in
employer’s home (%)
26 15
  
Table 8: Contraceptive use
N Percent
Have done something to avoid pregnancy, among girls
who have ever had sex
106 79
Used something to avoid getting pregnant last time had
sex
106 74
If used something to avoid pregnancy, method used:
Pill
Injectable/Depo‐Provera/Noristat
Diaphragm/foam tablets/jelly/cream
Male condom
Female condom
Safe days/periodic abstinence
Emergency contraception
Withdrawal
77
77
77
77
77
77
77
77
4
4
1
72
5
10
3
4
  
21
Table 9: Knowledge and use of RH and other services  
N Percent
Services sought out in last 12 months
VCT
FP
Legal Help Rape Crisis Center
Coalition on Violence Against Women  
165
165
165
163
57
4
3
2
Knows where to get tested for HIV 165 99
Knows the following places to get tested for HIV:  
Private doctor
Hospital  
Health Center
Private clinic
VCT Newstart center
Other
163
165
164
165
165
165
10
75
18
22
81
15
Has been tested for HIV 164 73
Has had symptoms of an STI in the last 12 months 166 7
Course of action if had STI symptoms:
Went to hospital
Nothing, it went away without treatment
12
12
67
33
  
22
Table 10: House girls’ Social Support
N Percent
Have
a special person to share joys and sorrows
Strongly disagree
Disagree
Agree
Strongly agree
166  9
13
32
46
Can talk to friends about problems
Strongly disagree
Disagree
Agree
Strongly agree  
166  
10
19
50
20
My employer cares about me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  7
12
47
34
<1
My employer respects me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  47
38
501
My employer’s children care about me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  7
11
40
31
11
My employer’s children respect me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  4
10
40
379
My employer’s spouse cares about me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  7
10
38
25
20
My employer’s spouse respects me
Strongly disagree
Disagree
Agree
166  73
39
23
Table 10: House girls’ Social Support
N Percent
Strongly agree
Not sure/ not applicable
36
16
My employer’s relatives care about me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  
10
20
38
22
10
My employer’s relatives respect me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  3
11
48
26
11
People in my church care about me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  58
43
33
10
People in my church respect me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  42
47
417
People in my community care about me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  8
19
43
228
People in my community respect me
Strongly disagree
Disagree
Agree
Strongly agree
Not sure/ not applicable
166  4
12
52
247
  
24
Table 11: House girls’ self‐efficacy in ability to choose sex and to seek out health care and
information  
Confidence in… N Percent
…ability to choose who have sex with
Not at all confident
Not confident
Confident
Very confident
Not sure
166  
2
13
24
60
<1
… to avoid sex any time  
Not at all confident
Not confident
Confident
Very confident
Not sure
166  
0.0
5
27
66
2
… to refuse to have sex if boyfriend will not use a condom
Not at all confident
Not confident
Confident
Very confident
Not sure
166  
<1
4
26
66
3
…to discuss contraceptives with a health care provider
Not confident
Confident
Very confident
Not sure
166  
2
37
59
2
… to discuss STIs/HIV/AIDS with a health care provider
Not confident
Confident
Very confident
Not sure
166  
<1
34
64
<1
… to visit a reproductive health care clinic
Not at all confident
Not confident
Confident
Very confident
Not sure
166  
<1
6
32
60
<1
… to visit a counselor if you were raped
Not at all confident
Not confident
Confident
Very confident
Not sure
166  
1
2
30
66
<1
… to attend a sex education class
Not at all confident
Not confident
166  
2
2
25
Table 11: House girls’ self‐efficacy in ability to choose sex and to seek out health care and
information  
Confidence in… N Percent
Confident
Very confident
30
67
Table 12: HELP Training
N Percent
Topics Covered During HELP Training:
Self‐ esteem and communications
Sexual and reproductive health
STIs including HIV prevention
Sexual violence, alcohol and drug abuse
Personal savings and financial management
Basic household security
Basic first aid
None
166
165
165
166
166
166
165
155
90
62
81
49
69
70
76
<1
Topics personally benefited from:
Self ‐esteem and communications
Sexual and reproductive health
STIs including HIV prevention
Sexual violence, alcohol and drug abuse
Personal savings and financial management
Basic household security
Basic first aid
None
Don’t know
Other
166
166
166
165
166
165
166
157
154
166
73
45
68
30
55
42
47
2
1
2
Least beneficial of the topics:
None 166 76
Table 13: Percent reporting changes in behaviors as a result of various training topics
N Percent
Are you doing anything different as a result of having
attending the training in…?
Self‐esteem and communications
STIs including HIV prevention
Sexual violence, alcohol and drug abuse
Sexual and reproductive health
Personal savings and financial management
Basic household security
Basic first aid
165
166
165
165
164
153
161
81
64
64
51
76
68
49
  
  
26
Table 14: Main reported behavior changes as a result of various training topics
Self‐esteem and communications
Has increased respect for self, employer and/or others
Communicates better (can say no, is more open, more polite)
Appreciates her work more/ is proud of being a house girl
More confident
Values herself more/ increased self‐esteem
STIs including HIV prevention
Plans to abstain/ abstain until marriage
Uses condoms/ protects herself from infection
Is being faithful
Tells her friends about STIs and HIV
Sexual violence, alcohol and drug abuse
Avoids dangerous/isolated places that can lead to rape
Avoids bad influences/ bad company/ peer pressure
Avoids drugs and alcohol
Sexual and reproductive health
Abstains
Uses condoms
Protects against unwanted pregnancy/ using family planning
Shares information with friends
Personal savings and financial management
Saves more money
Made a budget
Does not spend as much
Opened savings account/ saving through home bank
Basic household security
Keeps household doors and windows closed/ locked
Doesn’t open the door/ gate to strangers
Doesn’t leave house keys anywhere/ with anyone
Basic first aid
Does basic first aid (handles cuts, bleeding, fever…)
Is more careful in house
Knows what to do if someone chokes
Knows how to handle an emergency/ more confident
  
27
Appendix B: Results of the Formative 
Assessment 
The qualitative formative assessment was carried out in the absence of any detailed or reliable
information about the potential sources of vulnerability of house girls to STIs/HIV and unintended
pregnancies and about the feasibility of reaching them with interventions, or even assessments. The
results indicated that house girls are highly vulnerable, and that they are willing and able to participate
in interventions through the church. Below is a summary of those findings.
The results of the formative assessment with 18 house girls aged 18‐22 showed that house girls are a
particularly vulnerable group to STIs/HIV and unintended pregnancies, because of the combination of
their poor socioeconomic background, their isolation and lack of social support, the low status of their
work, and their experiences with sexual coercion and other forms of violence (Figure 1).  
Figure 1: Sources of Vulnerability for House Girls in Bahati, Nairobi
The 18 house girls with whom we conducted in‐depth interviews were primarily from large, poor, rural
families who migrated to Nairobi to support themselves and their families. Almost all stopped their
schooling after primary school for financial reasons. There is strong evidence to suggest a direct link
between women’s low economic status and their vulnerability and exposure to HIV (WHO, 2002;
Ziegler & Krieger, 1997). In particular, house girls’ poor socioeconomic situations increases their
vulnerability to HIV and unintended pregnancy because they may be more likely to fear a) losing their
job, b) letting their family down, c) destitution, and/or d) abandonment, which means that they may be
more likely to put up with abusive situations at work.  
Socio-economic
background
Experiences 
with 
Violence/ 
coercion
Low status of work
Isolation/ 
Lack of 
social 
support
28
The assessment also revealed that domestic work is considered a low status job in Kenya by both
house girls and society. This means that despite the enormous amount of work that house girls
perform for the families where they work – 18‐hour days being the norm – they are not paid a
commensurate salary, or even minimum wage8, and they are treated as inferior members of society.
The low status of domestic work may lead to a low (or lower) self‐esteem because of being
systematically treated by society and their immediate contacts as inferior. These factors may increase
house girls’ vulnerability to STIs/HIV because low social and economic status has also been associated
with less ability to negotiate condom use or to get out of abusive relationships (WHO, 2002).
House girls are isolated from their families and the rest of society at their places of work, and are
allowed little time off from work, which means that they have less access to social support than other
young Kenyan women. Twelve of the house girls worked more than 100 kilometers from their home of
origin. Several reported that when they came to Nairobi they did not know their way around and did
not have anyone to help them. They also reported that that when they have problems in the
households where they work they often have no one to turn to. Developing and accessing these
sources of social support is an important protective factor for risky behavior that is not available to
house girls (Springer et al, 2006; Ngom, 2003). Some house girls can ask for time off, under special
circumstances, but others only get time off for church on Sundays (and then they are often
accompanied by the children of the family).  
House girls’ isolation and poor socio‐economic background are also likely the causes of the lack of
information and low use of services that we saw among the house girls we interviewed. For example,
although most had had sex, many did not know what an STI was (except for HIV), and among those
who did, many could not name their symptoms. None had visited a voluntary counseling and testing
center.  
At the same time, the house girls we interviewed were not currently sexually active – only four of them
said they had had sex in the last six months. Most were either not interested or simply have no time.
But this does not mean that they are less vulnerable to future risky sex. On the contrary, it may be that
their isolation means that they are not gaining the negotiation skills that their peers are more likely to
be getting so that when they do initiate sex they will be more at risk. This seems to be substantiated by
those who are sexually active, several of whom are in risky sexual relationships, either because their
partner has multiple partners, or because they are not using condoms.
House girls are also vulnerable due to their previous experiences with violence and sexual abuse, both
outside and inside the workplace. In this assessment, eight of the house girls said they had experienced
unwanted advances by either a male member or a male friend of the employer’s family, which could
take the form of verbal harassment, inappropriate touching, or attempted rape. In addition, most of
the house girls had previous experiences of violence and sexual coercion. Multi‐country research has
found that experience of sexual coercion as an adolescent is associated with a greater likelihood of
experiencing multiple episodes of sexual coercion, risky sexual behaviors, adverse mental outcomes
(such as low self‐esteem, depression, anxiety), and a greater incidence of STIs, HIV, and unintended
pregnancy (Ganju et al, 2004). Thus, this group of women is prone to behaviors and abusive situations
that would place them at a higher risk of HIV and unintended pregnancies.
                                                      
8
The average monthly salary of the house girls we interviewed was KES 1,650/month (23 US$). The official level of minimum wage in 
Kenya is KES 2.500/month (34 US$). 
29
A final source of vulnerability for house girls is their age. In this assessment, we were unable to
interview house girls under 18 because of lack of access to parental consent and because we were
unwilling to ask employers for permission for fear of coercion. There are no reliable statistics on the
age of house girls in Kenya, but one publication estimated that over 90% were under 18 (RoK/Kenya,
1997). Domestic workers are amongst the most excluded and invisible children in the globally, and thus
some of the most vulnerable. The most recent State of the World’s Children defines exclusion and
invisibility in the following manner:
“Children are considered as excluded relative to other children if they are deemed at risk of
missing out on an environment that protects them from violence, abuse and exploitation, or if
they are unable to access essential services and goods in a way that threatens their ability to
participate fully in society in the future. Children may be excluded by their family, the
community, government, civil society, the media, the private sector and other children.”
(UNICEF, 2006, p. 7)
According to this definition, and taking into account the results above, adolescent house girls would
certainly be considered a particularly vulnerable group.
Feasibility
In order to determine the feasibility of an intervention with house girls to reduce their vulnerability to
STIs/HIV and unintended pregnancy we looked at five aspects: 1) house girls’ interest and availability,
2) employers’ interest and house girls availability from the employers’ perspective, 3) the types of
information house girls would like to receive, 4) the types of information that employers want their
house girls to receive, and 5) house girls’ literacy levels and access to media. To summarize, all house
girls were interested in participating in some kind of an intervention, but said they would need to get
permission from their employers. Employers also said they would have no problem with house girls
participating in an intervention to reduce their risk of HIV and unintended pregnancy, but they would
need more information to make sure that it was legitimate. However, they did not think that other
employers would agree, either because they do not see house girls as human beings or because they
would be afraid that the house girl would leave their employ. House girls said they were available on
Sundays, but employers generally preferred their house girls to take time off during weekdays when
their children were in school. In terms of training topics, house girls tended to mention improving their
health knowledge and joining together with other house girls for support and advocacy. Employers also
mentioned support and information for house girls, but also focused on concrete job skills such as
tailoring and baking. Almost all house girls said that they listen to a radio during the day. Some had
access to television or print media.
  
30
APPENDIX C: House Girls phase 2 Baseline 
Survey, Summary of findings 
INTRODUCTION
The President’s Emergency Fund for AIDS Relief allocated funds in 2007 for a multilayered church‐
based intervention with house girls and other members of the community in and around the Bahati
Estate in Nairobi with the overall goal of reaching house girls with HIV and unintended pregnancy
prevention messages. The award was based on results from a qualitative, formative assessment carried
out with 18 house girls aged 18‐22 by FHI 360 and Kenyatta University (KU) at the Presbyterian Church
of East Africa (P.C.E.A.) Bahati Martyrs Parish that identified house girls as a population particularly
vulnerable to HIV and unintended pregnancy. Specifically, the formative assessment revealed that
house girls are at significant risk of acquiring HIV and unintended pregnancies because of 1) their
socioeconomic background; 2) isolation and lack of social support; 3) the low status of their work; and
4) their previous experiences with sexual coercion and violence. Finally, the formative research
revealed that house girls have some knowledge of modes of transmission and prevention of HIV, but
knowledge and use of contraception and condoms are low9.
The formative assessment provided a contextual basis for the development of a program to reduce
house girls’ vulnerabilities, but the sample was small and it was limited to house girls aged 18‐22. Given
the likelihood of younger house girls being more vulnerable to all of the factors mentioned above, it
was important to collect information from a sample that also included younger women. Thus, using the
results of the formative assessment as a basis, a survey was planned targeting approximately 100
house girls aged 14‐22 in the PCEA Bahati Martyrs Parish.  
GOALS AND OBJECTIVES
This survey’s goal was to validate the findings in the formative assessment (Thomsen et al, 2007). By
expanding the sample to at least 100 using a primarily quantitative instrument, we would be able to
determine if the sources of vulnerability to HIV are shared in the larger house girl population in the
church. The survey was intended to provide a more representative sample of house girls from the
church in order to inform the program activities by interviewing all eligible house girls’ ages 14‐22
consenting to interviews. However, the youngest house girl who participated in the survey was 16, and
thus we were unable to collect data from younger house girls who may be the most vulnerable group.
The survey had the following primary and secondary objectives:
Primary objective:  
1) To assess house girls’ knowledge of HIV, sexually transmitted infections (STIs), and reproductive
health.
Secondary objectives:
2) To examine house girls’ sources of vulnerability to STI/ HIV and pregnancy.
3) To document house girls’ sexual behavior
4) To determine house girls’ capacity to take risk‐reduction action
                                                      
9
 More details on the results of the assessment are available in Appendix A. 
31
METHODS
The site for the survey was the PCEA Bahati Martyrs Parish in the Bahati estate of Nairobi which has
between 7‐8,000 congregants, making it one of the largest congregations in Nairobi. This is the same
site as that of the formative assessment. A cross‐sectional survey of all eligible and consenting house
girls aged 14‐22 in the congregation was conducted.  
The survey was administered to house girls whose employers attend the PCEA Bahati Martyrs Parish. In
some cases the house girls also attended the church, but this was not a requirement for participating in
the survey as long as the employer was a congregant. In addition, the participants needed to meet the
following criteria:  
1. Living with employer;
2. Not planning on moving in the next year;
3. Aged 14‐2210
The projected sample size was 100 house girls, with a provision to continue to include interested
eligible participants since sufficient resources were available. The final number interviewed was 153.
House girls who participated in the survey ranged in age from 16‐22. Because of the lack of younger
girls (aged 14‐15), results were not stratified by age.
RESULTS AND DISCUSSION
Socio‐demographic information
General socio‐demographic information for the 153 house girls shows they ranged in age from 16 to
22, with an average age of about 20 years. Only ten percent were under the age of 18. The youngest
house girl who participated in the survey was 16, and thus we were unable to collect data from
younger house girls who may be the most vulnerable group.  
More than half of the house girls had only received primary school education. About one‐quarter of
house girls reported being a member of a youth club or organization.
Data about house girls’ employment showed the average length of time a study participant had
worked as a house girl is 2 years and the average length of time with the current employer was 1 year.
Given the average age of the survey population as 20, this means that, on average, these house girls
started working at the age of 18.
Eighty‐seven percent of all house girls have part or all of Sunday off from their duties as a house girl.
This indicates that Sundays might be the best choice of time for an intervention or program designed
to assist house girls.
House girls’ knowledge of HIV, STIs, and reproductive health
Most house girls knew that a girl can get pregnant the first time she has sex. However, just under half
knew that a girl can get pregnant if she has sex while standing up. Around one‐tenth of the house girls
                                                      
10 It was important to include house girls under 18 because they are likely more vulnerable than their older counterparts. The upper age limit is set at 22 to include those who participated in the formative 
research study if they desire. 
32
knew that a girl can get pregnant if the boy withdraws before ejaculation. Less than 15% knew that a
woman has the greatest chance of becoming pregnant in the middle of her menstrual cycle. This lack of
knowledge about when a girl can get pregnant could make house girls vulnerable to unintended
pregnancies.
Knowledge of signs and symptoms of STIs was relatively low; the most commonly mentioned sign was
burning pain or itching in the genital area, mentioned by under 40 percent of house girls. Knowledge of
ways to avoid STIs was better, with more than half of the house girls knowing that abstinence and/or
condoms could prevent STIs. However, almost half of house girls did not spontaneously mention
condoms as a way to avoid getting STIs. House girls’ lack of knowledge about STIs could put them at
risk of being infected; in addition, their low knowledge of signs and symptoms of STIs could indicate
that some house girls are infected without realizing it.
More than half of the house girls surveyed knew that abstinence and/or condoms could prevent AIDS.
Avoiding needles and being faithful to one partner were other commonly mentioned prevention
methods. Over three‐quarters of the house girls knew that condoms could prevent pregnancy and STIs
including HIV. The generally stronger knowledge about HIV among house girls is potentially an
indication of the effectiveness of HIV information programs. However, whatever information house
girls are getting about HIV may be neglecting to inform them about the risks of STIs.
Ninety‐two percent of house girls named two or more advantages of using condoms. Eighty‐five
percent said that pregnancy prevention and HIV prevention were two advantages of using condoms.
Disadvantages of condom use were generally mentioned by fewer house girls. Around one‐quarter of
them said that condoms could burst. Fifteen percent said that condoms were unsafe or not 100 %
effective. Twelve percent said that there were no disadvantages of using condoms and 45% said they
didn’t know the disadvantages of using condoms.
House girls’ sources of vulnerability to STI/ HIV and pregnancy
The formative assessment revealed that house girls lack strong social support in the communities
where they work. This lack of support leaves them without someone to talk to or seek advice from, and
thus leaves them vulnerable to STIs including HIV and unintended pregnancy.
Table 1 shows the Sense of Belonging scale and score. Respondents were read a series of statements
about their social support and community appreciation and asked to strongly disagree, disagree, agree,
or strongly agree. There was no option for don’t know, unsure, or no answer. Answers were scored
from 1 for strongly disagree to 4 for strongly agree. The total score possible was 40 and one house girl
scored 40. The lowest possible score was 10; the lowest score recorded was 15. The average total
score was 30.5 out of 40.
33
Table 1: Sense of Belonging
Social Support
%
(n=153)
Strongly disagree
(1 pt)
Disagree  
(2 pts)
Agree  
(3 pts)
Strongly agree (4
pts)
There is a special person with whom
you can share joys and sorrows
12 10 27 51
You can talk about your problems
with your friends 17 14 39 30
Your employer cares about you 20 13 40 27
Your employer respects you 16 16 33 36
Your employer’s other family
members care about you 25 19 29 28
Your employer’s other family
members respect you 14 16 34 36
People in your church care about
you
6 5 25 64
People in your church respect you 2 1 27 70
Other community members care
about you 9 18 41 33
Other community members respect
you
5 8 39 49
Overall Sense of Belonging score
(max possible =40)
Lowest
Highest
Mean
15
40
30.5
Three‐fourths of the house girls agreed or strongly agreed that there is a special person with whom
they can share joys and sorrows. Approximately 90 percent of house girls either agreed or strongly
agreed that people in their church care about them and that people in their church respect them.
House girls may consider church to be a safe haven, where they are well‐treated. This is a further
indication that churches may be a good site for any outreach programs for house girls.  
For no statements did more than half of the house girls disagree or strongly disagree. The strongest
area of disagreement was with the statement “Your employer’s other family members care about
you,” with approximately 44% disagreeing or strongly disagreeing. “Other family members” is unclear
and could include the spouse of the employer, the employer’s children, or other members of the
extended family. House girls’ employers are generally the women in the family, so “other family
members” could include husbands.
Data on the house girls’ relationship and sexual history show that just over half of all house girls
reported that they had a boyfriend at the time of the survey. Of those, under half said that they had
had sex with their current boyfriend. Just over half of all house girls reported ever having sex, including
consensual, non‐consensual, and transactional sex. Of those, about 40 percent had only had one
34
partner in their lifetime, 36% had had two partners, and 22% said they had had two or more partners.
In the past year, most house girls with consensual sexual experience (65%) had one partner and about
one‐fourth said they had no partners in the past year.
Data about house girls’ experience with non‐consensual and transactional sex shows that of 80 house
girls who reported ever having sex, 12 percent (ten house girls) said they had ever received money,
gifts, or rewards for sex. Eight percent (six house girls) said they had ever been physically forced to
have sex; four of these house girls stated that they were physically forced to have sex in an employer’s
house. However, there is no information on who forced them to have sex within their employer’s
house. Ninety‐six percent of the 80 house girls said that they had ever had consensual sex, or sex that
they desired. The remaining three house girls reported only ever having unwanted or non‐consensual
sex.  
Contraceptive use by house girls who had sex in the past 12 months is shown in Table 2.
Table 2: Contraceptive Use by House Girls  
Who Had Sex in the Last 12 Months
Among those who had sex in the past 12 months, more than 80 percent used a contraceptive method.
The most commonly used method was a condom. Safe days or periodic abstinence was the second
most commonly used method. Only 16 percent of these house girls had used a modern, hormonal
method of contraception, including the pill and implants.
Seven house girls reported using two methods in the last 12 months, although it is impossible to know
if these methods were used for the same sex acts. Two house girls said they used both pills and
condoms; two said they used both emergency contraception and condoms; two said they used safe
days/periodic abstinence and condoms; and one reported using both implants and condoms (data not
shown).
For the ten sexually active house girls who did not use a method in the last 12 months, the most
common reason given was that they did not know what contraception was, mentioned by three house
girls. Two said that their partner did not like contraceptives. One twenty‐one‐year‐old house girl said
she wanted to get pregnant and one girl said that she did not use a method because the sex was
forced.  Table 3 shows other information on the reproductive health history of house girls.
Contraceptive Use %
Used any contraceptive methods in the last 12 months
Yes
No
(n=59)
83
17
Methods used*
Condom
Safe days/periodic abstinence
Pill
Norplant/Implants
Emergency contraception
Injectable/Depo‐Provera/Noristerat
(n=49)
78
16
10
4
4
2
*Multiple responses are possible.
35
Table 3: Reproductive Health History
Reproductive Health History %
Partner has ever used male condom with her
Yes
No
Don’t know/Don’t remember
(n=77)*
75
23
1
Any STI symptoms in last 12 months
Yes
No
No answer
(n=77)*
6
92
1
Ever been pregnant
Yes
No
(n=77)*
27
73
Ever tested for HIV
Yes
No
(n=153)†
49
51
Ever tested for HIV (only those who ever had sex)
Yes
No
(n=80)‡
64
36
*Among house girls who reported ever having consensual sex
†
Among all house girls
‡ Among house girls who reported ever having sex
Among the 77 house girls who have ever had consensual sex, three‐quarters said that a partner had
used a male condom with her. Only five reported having symptoms of an STI in the past year; all of
these house girls said they sought treatment for those symptoms.
About one‐quarter of house girls who had ever had consensual sex reported ever being pregnant.
These girls were not asked about the outcomes of their pregnancies.
Data on HIV testing is shown for all house girls; just under half reported being tested for HIV. For the
80 house girls who ever reported having sex, including the three house girls who had only ever had
non‐consensual sex, 64% said they had been tested for HIV. About one‐third of the house girls who
reported never having sex said they had been tested for HIV (data not shown). House girls were not
asked to share the results of their HIV test.
House girls’ capacity to take risk‐reduction action
House girls’ capacity to reduce their risk of acquiring HIV/AIDS or unwanted pregnancy is related to the
personal and social resources to which they have access. This includes their access to media and
information about HIV/AIDS. It also includes their confidence in their own ability to make certain
choices or take specific actions.
Table 4 shows information about house girls’ access to media.
36
Table 4: Access to Information
Access to Information %
Frequency of reading a newspaper or magazine
Every day or almost every day
At least once a week
Once a month or less often
Never
No answer
(n=153)
10
41
23
25
1
Newspapers or magazines read*
Daily Nation Newspaper
Parents Magazine
Sunday Nation Newspaper
East African Standard Newspaper
Taifa Leo Newspaper
(n=113)
57
46
12
11
11
Frequency of listening to the radio
Every day or almost every day
Once a week or less often
Never
(n=153)
84
10
6
Radio stations listened to*
Kameme FM
Kiss FM
Inooro FM
Easy FM
Metro FM
Citizen FM
Cooro FM
(n=144)
62
41
28
19
18
18
16
Frequency of watching TV
Every day or almost every day
At least once a week
Never
(n=153)
63
22
14
*Multiple responses are possible; only selected items shown.
The most commonly read magazines were the Daily Nation Newspaper and Parents Magazine. About
half of house girls said they read a newspaper or magazine at least once a week. The radio stations
most commonly listened to by house girls at the time of survey were Kameme FM and Kiss FM.
Overall, 90 percent of the house girls said they read a newspaper or magazine, listened to the radio,
and/or watched TV every day or almost every day. Nine percent said they accessed at least one of
these forms of media weekly; one house girl said she accessed at least one of these forms of media
monthly. Two house girls reported no regular access to any of these media. These data show that
house girls have regular access to different forms of media, where they could hear HIV/AIDS or other
important messages.
House girls’ sources of information about HIV/AIDS are shown in Table 5; the data are presented in
several categories. Media sources were mentioned the most frequently, with radio and television
37
mentioned the most often overall. This corresponds with house girls’ regular access to media,
particularly radio and television.
Table 5: Information about HIV/AIDS
Information about HIV/AIDS
%
(n=153)
Sources of information on HIV/AIDS in past 12 months*
Media
Radio
Television
Newspaper
People  
Someone at church
Friends (not house girls)
Health care provider
Other house girls
Relative
Employer
Other
Course at Bahati Martyrs PCEA Church
School
None
73
50
20
41
21
20
17
14
11
11
10
1
*Multiple responses are possible; only selected items shown.
Among people mentioned, someone at church was mentioned by about 40 percent of house girls.
Friends (not house girls), health care providers, and other house girls were all mentioned by around
one‐fifth of house girls as sources of information.
Only two house girls said they had no source of information about HIV/AIDS; both of these house girls
said they listened to the radio and watched television every day or almost every day.
House girls’ access to health and social services is displayed in Table 6.
Table 6: Access to services
Access to services %
Services sought in past 12 months*
VCT
Family planning
Legal help
Rape crisis center
Coalition on Violence against Women (COVAW)
None
(n=153)
30
7
2
1
3
66
38
Thirty percent of the house girls had sought VCT services in the past 12 months; this included 30 house
girls who reported ever having sex and 16 house girls who said they had never had sex. Other services
sought were family planning, legal help, and the Coalition on Violence against Women (COVAW). One
house girl, who said she had been raped in the employer’s home, sought services at a rape crisis
center.
Sixty‐six percent of house girls said they had sought no services in the past 12 months. House girls got
advice on the services sought from various people, including relatives, friends, and health care
providers.
All but seven of the 153 house girls knew where to go for an HIV test. Among those who knew where
to go for an HIV test, more than 80 percent said HIV tests were available at public/government
facilities and/or VCT/Newstart centers.
Self‐efficacy questions indicated that three‐quarters of the house girls surveyed definitely could discuss
STIs/HIV with a health care provider, visit a counselor if raped, and/or attend a sex education class.
Two‐thirds of the house girls said they definitely could discuss contraceptives with a health care
provider. Over half of the house girls said they definitely could avoid sex any time they didn’t want it,
refuse to have sex if their boyfriend will not use a condom, and/or visit a RH clinic.
About one‐quarter of the house girls said they definitely could not refuse to have sex if their boyfriend
would not use a condom. In addition, about one‐quarter of the house girls said they either definitely
could not or probably could not avoid sex any time they didn’t want it.
It should be noted that the highest levels of confidence involve matters not directly related to sex. For
issues involving sex, the percent of house girls who responded “Definitely could” drops from the 60s to
around 50 percent. These data reveal house girls’ potential vulnerability to non‐consensual or
unprotected sex.  
Overall vulnerability
Source of advice on services sought in past 12 months*
A relative
Friends (not house girls)
Health care provider
Nobody
Other house girls
Employer
(n=53)
25
23
17
13
9
9
Knows where to go for an HIV test
Yes
No
(n=153)
95
5
Where to go for an HIV test*
Public/Government Health Facility
VCT/Newstart Center
Private Health Facility
(n=146)
84
83
46
*Multiple responses are possible; only selected items are shown.
39
To evaluate house girls’ overall vulnerability, a composite score was created, incorporating items from
the tables presented above. Because only about half of the house girls reported ever having sex, the
list of score elements was expanded beyond sexual experience to include items from the Sense of
Belonging and Self Efficacy scores.
The risk score included the following items:
Sexual behavior:
 Multiple partners in the last 12 months
 Ever have had transactional sex
 Ever physically forced to have sex
 STI symptoms in last 12 months
Self‐Efficacy:
 Definitely could not choose sexual partner
 Definitely could not avoid sex
 Definitely could not refuse sex with a boyfriend who won’t use a condom
Sense of Belonging:
 Strongly disagree about having a special person to talk to
 Strongly disagree about being able to talk about problems with friends
The goal of the composite vulnerability score was to determine how many house girls had more than
one of these risk factors listed above. The score was not weighted, but each item was given one point,
with a total possible score of 9. Results from the composite score are shown in Table 7.  
Table 7: Composite Vulnerability Score
Score
%
(n=153)
0
1
2
3
4
5
42
32
16
6
2
2
Mean score 1.0
Forty‐two percent of house girls had none of the above risk factors. Thirty‐two percent only had one.
One‐quarter of the house girls had two or more risk factors; this group is the group we examined more
closely to see which risk factors co‐occurred.
Among these house girls with two or more risk factors are all six house girls who reported ever being
forced to have sex, eight of the ten house girls who said they had ever exchanged sex for money or
gifts, and four of the five who reported having symptoms of an STI in the past year. No house girl
reported all three risk factors of rape, transactional sex, and STI symptoms, although six house girls
reported two out of the three. Only one of the three house girls who reported only ever having non‐
40
consensual sex had more than one risk factor; the other two scored very well on both the Sense of
Belonging and the Self‐Efficacy scales.
Twelve house girls who reported never having sex had more than one risk factor. For each of these
girls, their only risk factors came from the two scales, with an emphasis on the three factors from the
Self‐Efficacy scale. These house girls could benefit greatly from an intervention that would include
esteem building or empowerment exercises.
House girls in the high risk category (two or more risk factors) were compared to those with no risk
factors or only one risk factor. There were no statistically significant differences in terms of age, having
any time off work, education, access to media, or average monthly income. This means that it may be
very difficult to target these vulnerable house girls for interventions or programs.
CONCLUSIONS
This survey verified the findings of the formative assessment and provided a greater understanding of
the knowledge of STI/HIV/AIDS and behavior of house girls in the PCEA Bahati Martyrs Parish. The
survey results also reveal the extent and various causes of this population’s vulnerability to HIV,
suggesting potential linkages to risk‐reduction interventions.
House girls’ have poor to fair knowledge of issues concerning reproductive health and STIs/HIV/AIDS.
Of particular concern is the fact that under 15 percent of house girls surveyed knew the time during
the menstrual cycle when a woman can get pregnant. Also of concern is that house girls had very poor
knowledge of symptoms of STIs.
One of the results that could be of most utility to future interventions was condom use. Knowledge of
condoms was good, with the condom being cited as a way to avoid pregnancy, a way to avoid STIs, a
way to avoid HIV/AIDS, and as the method that will prevent both pregnancy and STIs/HIV/AIDS. In
addition, more house girls cited advantages of condom use than disadvantages. Also, three‐quarters of
the house girls who had ever had consensual sex reported using a condom with their partner at least
once. However, over a third of the house girls said they either "definitely could not" or "probably could
not" refuse to have sex if their boyfriend refused to have a condom. This indicates that house girls
would benefit from counseling on negotiating condom use.  
Awareness of some modern contraceptive methods was good, especially for pills (71%) and the
injectable (53%). However, among those who had heard of the different modern methods, only about
half could describe how the method works. In addition, use of modern contraceptive methods was low,
with only a small number saying that they had used the pill, the injectable, or Norplant in the past 12
months. Most house girls who had used a method to prevent pregnancy in the last 12 months said they
had used condoms; safe days or periodic abstinence was the second most common method, and the
pill was third. Four of the ten house girls who said they used no method to avoid pregnancy in the last
12 months indicated that lack of knowledge of methods or of how to use methods was the main
reason they did not use one.  
House girls would benefit from education or counseling on reproductive health and STIs and HIV/AIDS.
For reproductive health, topics should include basics like the menstrual cycle, modern methods of
contraception, and how those methods work to prevent pregnancy. Only a few house girls had sought
FP services in the past 12 months, so any education on reproductive health should include information
on where and how to obtain modern contraceptive methods.
41
Education or counseling on STIs and HIV/AIDS should include the symptoms of STIs, so that house girls
will know when they need to seek treatment. There should also be an emphasis on ways to avoid STIs
and HIV/AIDS, since a sizeable majority did not know the best ways to avoid these infections. It is
important that this education include both STIs and HIV/AIDS, as house girls’ knowledge of STIs is
especially lacking.
Other information to emphasize in education or counseling for house girls should include financial
training, specifically in the area of saving. This would also help increase the girls’ self‐esteem and give
them confidence to negotiate safer sexual health practices.
Most house girls have some time off during the week, with some time on Sundays being the most
common among our population in the PCEA Bahati Martyrs Parish. Thus, an intervention to target
house girls should be scheduled on Sundays, or whatever the community Sabbath or prayer day is, in
order to reach the maximum number of house girls. In addition, most house girls have access to media
on a daily or almost daily basis. An intervention through the media, targeting some popular
newspapers, magazines, radio stations, and TV channels, could reach house girls.
One‐quarter of the house girls surveyed could be considered highly vulnerable, having more than one
selected risk factors. This sub‐set of house girls is indistinguishable from the rest of our study
population in terms of socio‐demographic information, so it may be difficult to target them specifically.
However, they are also not different from other house girls in terms of access to media or time off
from work, so any intervention to target house girls through the media or during expected time off
would reach this group at higher risk.
It is also important to remember that house girls are a special out of school youth group who are not
considered and do not consider themselves part of the mainstream youth for whom HIV prevention
messages are developed. Building house girls’ self‐esteem and self‐worth could help them appreciate
their own value and not as societal misfits. This could make them more open to receiving broader HIV
prevention messages.
This study was intended to interview house girls aged 14‐22, but only house girls aged 16‐22
volunteered, with only a handful being under the age of 18. Another potential limitation to this survey
was that house girls who were interviewed were those with time off on Sundays so they could attend
church. It is probable that the most vulnerable house girls, including those under the age of 16 and
those who are unable to get time off to attend church, were not reached by this study. This is still a
population in need of attention.
  
© 42 2013