The evaluation of the JEWEL project: An innovative economic
enhancement and HIV prevention intervention study targeting drug
using women involved in prostitution
S. G. SHERMAN, D. GERMAN, Y. CHENG, M. MARKS, & M. BAILEY-KLOCHE
Johns Hopkins Bloomberg School of Public Health, Baltimore
The JEWEL (Jewellery Education for Women Empowering Their Lives) pilot study examined the efficacy of an economic
empowerment and HIV prevention intervention targeting illicit drug-using women (n? /50) who were involved in
prostitution in Baltimore, Maryland. The intervention was comprised of six 2-hour sessions that taught HIV prevention
risk reduction and the making, marketing and selling of jewellery. Bivariate comparisons examined behaviour change pre-
and 3-months post-intervention. The intervention’s effect on the change in the number of sex trade partners from baseline
to follow-up was explored with multiple linear regression. Participants were 62.0% African American, 5.0% were currently
employed, and the median age was 39 years old (Inter Quartile Range [IQR]: 34?/45). Women attended an average of six
(IQR: 4.5?/6.0) sessions. The women sold over $7,000 worth of jewellery in eleven sales. In comparing self-reported risk
behaviours pre and 3-month post intervention participation, we found significant reductions in: receiving drugs or money
for sex (100% versus 71.0%, p B /0.0005); the median number of sex trade partners per month (9 versus 3, p? /0.02); daily
drug use (76.0% vs. 55.0%, p? /0.003); the amount of money spent on drugs daily (US$52.57 versus US$46.71, p? /0.01);
and daily crack use (27.3% versus 13.1.0%, p? /0.014). In the presence of other variables in a multivariate linear model,
income from the jewelry sale was associated with a reduction in the number of sex trade partners at follow-up. The pilot
indicated effectiveness of a novel, HIV prevention, economic enhancement intervention upon HIV sexual risk behaviours
and drug utilization patterns.
The preponderance of epidemiological research on
female drug users’ risk of HIV focuses on individual-
level sociodemographic factors. A number of studies
have documented elevated sexual risk behaviours
among drug using women compared to their male
counterparts, resulting in significantly higher rates of
HIV and sexually transmitted infections (STIs)
(Doherty et al., 2000; Kral et al., 2000; Latka et
al., 2001; Sherman et al., 2003). Much of the
research examining the role of sexual risk as related
to HIVand STIs among women drug users has been
among crack smokers. Crack use has been associated
with increased sexual activity due to heightened
sexual arousal (Ross et al., 2002; Ross et al.,
1999), trading sex for money or drugs (Baseman et
al., 1999; Chiasson et al., 1991; Ross et al., 2002),
and having multiple sex partners (Ross et al., 2002).
While these are significant factors in HIV risk, such
an individual focus de-emphasizes the role of
broader relationship and structural factors that
shape women’s risk through multiple direct and
indirect pathways. The current study examines the
effects of a novel economic empowerment HIV
prevention intervention upon sexual and drug-re-
lated risk behaviours.
O’Leary (2001) described women’s structural
vulnerabilities as a ‘nexus of risk’, whereby interac-
tions between diverse social disadvantages facilitate
HIV risk-taking behaviours. Research points to
economic disadvantage as a crucial contextual HIV
risk factor among women (Klein et al., 2002;
O’Leary, 2001). Regardless of route of administra-
tion (i.e. injecting, snorting, or smoking), similar
social, cultural and economic, contextual factors
shape sexual and drug use behaviours among female
drug users (Rhodes & Quirk, 1998; Sanders-Phillips,
2002). Poverty is the central factor that characterizes
(Farmer, 1996; Fenton, 2004; Friedman, 1998;
Klein et al., 2002) which, on an individual level,
manifests as lack of access to and control over
actions, resources, and self-determination (Ulin,
Female drug users often live in the context of
severe economic disadvantage, fluctuating between
Correspondence: Susan G. Sherman, PhD, MPH, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, 615
N. Wolfe Street, E6006, Baltimore, MD, 21205, USA. Tel: (410) 614-3518. Fax: (410) 955-1383. E-mail: email@example.com
AIDS Care, January 2006; 18(1): 1?/11
ISSN 0954-0121 print/ISSN 1360-0451 online # 2006 Taylor & Francis
intermittent employment and unemployment (Zanis
et al., 1994). Their ‘work’ is often limited to illicit
low status ‘employment’ in the drug economy and
trading sex for money, drugs, or shelter (Connors,
1996; Ross et al., 2002). Women drug users may
become involved in the street-level sex industry
because of a complex array of socioeconomic factors
including low levels of education and limited gainful
employment skills coupled with financial need that
includes supporting a drug habit (Weeks et al.,
1998). Economically motivated sexual behaviours,
such as exchanging sex for money or drugs, com-
promises women’s ability to employ safer sex strate-
gies and challenges their ability to be concerned
about the long-term consequences of their sexual
behaviours. In turn, their risk of becoming infected
with HIV and other STIs is increased (Nelson et al.,
1995; Worth, 1990).
Although many researchers have identified a need
for structural and environmental interventions to
reduce HIV risks among drug users (Friedman,
2000; Miller & Neaigus, 2001; O’Leary, 2001;
Rhodes & Quirk, 1998; Sweat & Denison, 1995),
very few studies have attempted to do so. The
preponderance of HIV intervention research remains
focused on individual-level behaviour change by
employing rational decision-making theories such
as the health belief model (Fishbein & Ajzen, 1975),
the theory of reasoned action (Montana et al.,
1997), and social learning theory (Bandura, 1977).
While such theories are necessary in the develop-
ment of behavioural interventions, interventions
based solely on these theories have shown short-
term behaviour change efficacy in various high-risk
populations (DeHovitz et al., 1994; Kelly et al.,
1989; Susser et al., 1998) but have been less
successful in creating sustainable (more than six
months) behaviour change among drug users (Des
Jarlais et al., 1991; Oakley et al., 1995) and are
insufficient when offered alone (Latkin & Knowlton,
2000). Further, application of these theories among
drug users has been critiqued for their assumption of
behaviour as volitional, their reliance upon rational
thinking, and insufficient accounting for other fac-
tors that are socially, culturally, and economically
mediated (Rhodes et al., 1996).
It is well established that poverty is associated with
HIV risk behaviours and long-term drug use, but
there are few examples of HIV prevention interven-
tions among drug users that focus on expanding
their economic opportunities. Job skills training
could provide multiple benefits in the lives of drug
using women who are involved in the sex trade.
Though environmental interventions for HIV pre-
vention as defined by Sweat and Denison (1995)
have been slow to develop in the US, lessons can be
learned from other fields. The notion of economic
empowerment and job skills training for low-income
populations is well established in the economic
development literature, with numerous evaluations
of the impact of micro-enterprise industries on
women’s economic well being (Hashemi et al.,
1996; Schuler et al., 1997; Sebstad & Chen,
1996). Substantial research has been conducted in
the developing world on the impact of micro-
enterprise on women’s economic well being (Ha-
shemi et al., 1996; Schuler et al., 1997; Sebstad &
Chen, 1996). Throughout Asia and Africa, models
such as the Grameen Bank have been shown to
increase women’s economic well being, enhance
contraceptive use, and strengthen their position in
families (Hashemi et al., 1996; Schuler et al., 1997).
A small number of US studies focus on job
training for former drug users, with the explicit
goal of promoting long-term abstinence through
program participation contingent upon sobriety.
Evaluations of vocational interventions in differing
treatment modalities have demonstrated short-term
success but mixed results in long-term effects in
relapse prevention and increasing employment with
mixed results in long-term (Higgins & Silverman,
1999; Milby et al., 2003; Silverman et al., 2002). A
novel approach in this field is that of the therapeutic
work place, in which rewards for job training and
employment are contingent upon abstinence (Bige-
low & Silverman, 1999; Milby et al., 2003; Silver-
man et al., 2002; Widman et al., 2000). These
studies have shown mixed results in long-term
abstinence. Such highly behaviourally controlled
environments only reach the small percentage of
treatment seeking, drug users who are able to
comply with such a highly regulated environment
(Schottenfeld et al., 1992; Zanis et al., 1994).
Vocational training programmes have not been
tested among active drug users and as a result, there
is a dearth of information on how such programmes
affect illicit drug utilization patterns and HIV/STI
The current study describes results from a pilot
economic enhancement HIV prevention interven-
tion study upon sexual and drug-related risk beha-
viours among women drug users involved in the sex
trade. Economic enhancement was the broader
framework of the study in that it was designed to
address environmental barriers to prevention such as
lack of access to sustainable and licit income (Beeker
et al., 1998; Zimmerman, 2000; Zimmerman, 1995;
Brown, 1991). The pilot study had three primary
aims: 1) to assess the feasibility of training drug
using women how to make, market, and sell beaded
jewellery; 2) to enhance women’s HIV transmission
knowledge, provide risk, and critical thinking skills;
and 3) to enhance women’s job self efficacy by
S. G. Sherman et al.
providing them training in the making, marketing
and selling of beaded jewellery.
The pre- and 3-month post-intervention study was
conducted in Baltimore, Maryland. Between De-
cember, 2002 and June, 2003, 54 women were
recruited into the Jewellery and Education for
Women Empowering their Lives (JEWEL) study.
The study inclusion criteria were: being female;
being 18 to 45 years of age, having traded sex for
money or drugs in the past month; and having used
heroin and/or cocaine at least weekly in the past
month. If women were eligible for the study and
agreed to consent, they were administered a beha-
vioural survey and locator information was collected
to inform them of the start date of the intervention.
In total, 54 women participated in six cohorts of the
six-session intervention. Follow-up interviews were
conducted at three months after completion of the
intervention. Four women were ineligible for follow-
up: two were incarcerated; one was hospitalized; and
one was in a long-term drug treatment facility. Fifty
women received the follow-up survey, resulting in a
follow-up rate of 90%.
Of the 98 women screened, 77 (79%) women were
eligible and of those, 55 (70.0%) women attended
the study’s first intervention session. Of those, 50
(91%) returned for their 3-month follow-up survey.
The first-session attendance rate is acceptable in
light of those reported by large randomized beha-
vioural interventions (Latkin et al., 2002; NIMH,
Potential study participants were recruited through
targeted outreach in a neighbourhood with high rates
of drug use and street prostitution. Additionally, the
study was publicized at a nonprofit organization in
the target neighbourhood that provides services to
women involved in prostitution and at the Baltimore
Syringe Exchange Program van which was parked in
the neighbourhood twice a week. Women were
approached by trained study recruiters who had
extensive experience working with the target popula-
tion in previous studies. After women heard a brief
description of the study regarding a six session
intervention training women in HIV prevention
and jewellery making, they were asked to call or
drop by the study house for screening. The study
was housed in a store-front office located on a busy
street in the target neighbourhood.
The JEWEL intervention focused on HIV preven-
tion education as well as teaching the making,
marketing and selling of beaded jewellery. Jewellery
making was selected as the study’s economic em-
powerment activity because of its broad market
appeal, its relatively inexpensive start-up costs, ease
of learning, and the short amount of time needed to
produce marketable products. The intervention was
comprised of six 2-hour sessions that were divided
between HIV prevention and jewellery making. A
draft manual was piloted with 10 women from the
study population, which informed the finalized
version used throughout the study. The final manual
is available from the first author upon request.
The HIV prevention component was based on the
social cognitive theory (SCT), which is one of the
most widely used behavioural theory utilized in HIV
prevention research targeting drug users (Van Em-
pelen et al., 2003). SCT postulates that an indivi-
dual’s agency in behaviour change is, in part, shaped
and constrained by the socio-structural environ-
ment, such as economic conditions and life structure
(Bandura, 2001; Bandura, 1989; Bandura, 1986).
The intervention was highly structured and was led
by a trained female facilitator who followed the
scripted intervention manual. Many of the HIV
prevention activities were built on those that we
have utilized in our previous intervention research
(Latkin, et al., 2002). HIV prevention was delivered
through a variety of fun and interactive activities that
did not require high levels of literacy, such as games,
facilitated discussions, role-plays, and interactive
demonstrations. The range of activities was designed
to address different learning styles and to be
The HIV prevention components were based on
five elements in accordance with the social cognitive
theory: 1) to increase women’s HIV/STI disease and
risk reduction knowledge as well as the effects of
drugs in their lives; 2) to enhance women’s self-
efficacy in practicing safer sex with specific partner
types and increase their outcome expectations; 3)
to teach participants concrete skills-building to
improve their efficacy in reducing sexual risk, risky
syringe-sharing behaviours, and drug use; 4) to teach
women sexual negotiation and communication skills;
and 5) to provide women with the opportunity to
practice newly acquired skills through role play and
activities. The jewellery-making component not only
taught delineated skills but also aimed to increase
their self-efficacy in entering job training pro-
grammes or seeking employment. The connection
between the jewellery making and risk reduction
components will be made through informal, facil-
itator-led conversations in the jewellery-making half
The JEWEL project
of the sessions. The HIV prevention content in
specific sessions was as follows: Session 1 ?/ drug
related risk reduction; Session 2 ?/ sexual risk
reduction in the context of primary, casual and trade
sexual partners and related role pays; Session 3 ?/
male and female condom demonstration, practice,
and related role plays; and Session 4 ?/ the connec-
tion between drug use and sex.
The jewellery making lessons were conducted by a
local artist with extensive experience in making
beaded jewellery. The jewellery making modules
became increasingly complex over the course of the
intervention. In Session 1, participants learned
about colours, textures, and materials in preparation
for making their first piece of jewellery, a bracelet. In
the following four sessions, they learned the neces-
sary techniques for making earrings and necklaces,
which varied in difficulty depending on the indivi-
dual’s ability and interest. In preparation for the
Session 5 jewellery sale, Session 4 focused on
marketing skills and included role plays with facil-
itators to practice how to market jewellery to
potential buyers. Additionally, women produced
‘risk reduction’ cards to accompany their individual
pieces of jewellery at the upcoming sale. The cards
contained messages about what the programme
meant to each woman. Session 5 was a jewellery
sale which occurred at the Johns Hopkins Outpatient
Center. The proceeds from the sales were divided as
follows in order to best reflect a ‘real world’ market:
50% to the individual jeweller; 30% divided between
participants who worked at the sale; and 20% for
The sixth session consisted of a review of all HIV
and risk reduction topics from the first four sessions,
focused discussion on local job-training and GED
programmes, and a graduation ceremony. Graduates
and current intervention participants also had the
opportunity to attend open jewellery making ses-
sions on a drop-in basis every Wednesday during the
course of the study. Jewellery from these sessions was
also sold during project sales with the same guide-
lines for distribution of proceeds.
The study was a pre- and three-month post-inter-
vention study design. Through the consent process,
an interviewer explained the study’s purpose and
issues related to confidentiality. Upon providing
written informed consent, women were administered
a baseline survey by a trained interviewer. Once
twelve women were administered the survey, a new
cohort began. Six cohorts ran consecutively from
January 2003 through June 2003. The cohorts met
twice weekly over a three-week period. Food was
provided during each session. Women completed a
follow-up interview three months after the last
session of their respective cohort. This survey
contained behavioural questions that were parallel
to those in the baseline survey in order to measure
change over time. Women were reimbursed $20 for
each interview and each intervention session as well
as a portion of the jewel sale proceeds (discussed
below). The study was approved by the Johns
Hopkins Bloomberg School of Public Health Com-
mittee on Human Research.
The baseline and three-month follow-up surveys
sources of income; current and past drug use
patterns; an extensive employment and job training
history; a job self efficacy scale; sexual behaviours
with primary, casual, and sex trade partners; and a
history of sexually transmitted infections. In both
surveys, all behavioural questions were asked in
reference to the last three months prior to interview.
Independent variables included the following demo-
graphic characteristics ascertained at baseline: med-
ian age; race; high school graduate status; recent
(past three months) homelessness; recent employ-
ment; ever participated in a job training programme;
ever and recently been incarcerated, having and
living with children; ever tested for HIV; and self-
reported HIV status. Intervention-related variables
that were examined included the median number of
sessions attended, and above and below the median
value of US$112.80 of money earned by those (n? /
42) who earned any income from jewellery sales.
All licit and illicit sources of income reported at
baseline and 3-month follow-up were examined. A
12-item job self-efficacy index was developed to
measure basic functions needed to maintain a job.
The scale’s internal consistency reliability was mea-
sured using Cronbach’s alpha. Regarding the struc-
ture of the set, all items correlated positively. All
seven items had correlations above the recom-
mended value of 0.20 (Carmines & Zeller, 1978).
The items were measured on a four-point Likert
scale (1? /not certain at all; 2? /not very certain;
3? /certain; 4? /very certain). The corresponding
alpha for the baseline index was 0.82 and the
follow-up index’s alpha was 0.78. The seven-item
scale was comprised of the following items: 1) How
certain are you that you could continue to work for
an entire morning even if you need a hit (of drugs)?;
2) How certain are you that you can be on time to a
job even if you are tired?; 3) How certain are you that
you can be on time to a job even if you are
S. G. Sherman et al.
withdrawing and feeling sick?; 4) How certain are
you that you could show up to work after missing a
day without calling?; 5) How certain are you that you
could do what someone tells you to do at a job?; 6)
How certain are you that you can ask for clarification
if someone tells you to do something that you don’t
understand?; and 7) How certain are you that you
can concentrate for several hours on work if it was
boring? Additionally, principal components analysis
demonstrated that the seven items were a part of a
one-factor solution by examination of the resulting
screen plot and Kaiser Guttman standards (Kaiser,
Drug utilization patterns at baseline and three-
month follow-up were explored, including: the use
and frequency of heroin, crack, and/or cocaine use
that was categorized above and below the median
value of daily drug use; the amount of money spent
on drugs on a typical day. Sexual behaviours
explored at baseline and three-month follow-up
were examined, including: number of sex partners;
types of sexual partners (main, casual and trade);
frequency of protected and unprotected anal, vaginal
and oral sex with each sexual partner type. Con-
tinuous variables such as amount of money spent of
drugs daily were divided around the median value.
Response categories to the categorical drug use and
sexual behavioural variables were reduced according
to distribution of the data (i.e. daily vs less than daily
(i.e. always used condom with specific sexual partner
The first series of analyses compared drug utilization
patterns, sexual risk behaviours, and sources of
income at baseline and three-months following the
intervention. Chi-square tests were used to compare
dichotomous and categorical data and paired t-tests
for matched observations were used to analyse
Unadjusted and adjusted multiple linear regres-
sion was employed to examine correlates of the
change in the number of sex trader partners from
baseline to follow-up. Independent variables found
to be significant below the 0.20 level in univariate
analyses or those that were hypothesized to be
potentially associated with the outcome was entered
into the exploratory multivariate regression models.
We examined the nature of the relationship between
continuous independent variables and the outcome.
Independent variables included in the model had a
linear relationship with the outcome. In building the
multiple model, variables were selected manually
with a back and forth step process and the final
model included variables that were scientifically and
statistically (B /0.10) significant. The current ana-
lyses were restricted to the 50 women who com-
pleted the three-month follow-up survey. Due to the
small sample size, we conducted simulation analyses
with 75 and 100 people to evaluate the power of our
study based on variance/covariance matrix from the
raw data (Wuensch, 2002).
Sociodemographic and intervention-related charac-
teristics are displayed in Table I. Participants were
62.0% African American, 33% had completed high
school or received the high school equivalency
degree (GED), and the median age was 39 years
old (Inter Quartile Range [IQR]: 34, 45). Twenty-
seven percent reported being homeless, 5.0% re-
ported having licit employment in the three months
prior to being interviewed, and 20.0% of the sample
reported having ever participated in a job training
programme. Regarding incarceration, 81.0% of the
sample had ever been in jail and 42% had been jailed
in the three months prior to being surveyed.
Seventy-six percent of the sample reported having
children, with only 11.0% living with at least one of
the children. All but one participant had been tested
for HIV, and 9.3% reported having a positive result
at baseline. No incident HIV cases were reported at
Participants attended a median of six intervention
sessions (IQR: 4.5, 6.0). In total, twenty women
Table I. Sociodemographic characteristics of JEWEL participants
at baseline, n? /50.
High school graduate or GED
Median age (IQR)
Ever participated in a job training programme
last three months
Have Children live with children
Tested for HIV HIV? /
Median number of intervention sessions
Median (IQR) amount earned from jewellery
39 (34, 45)
* Past three months.
The JEWEL project
attended at least one open jewellery making session
and there was an average of five participants per
session (not shown). Net jewellery sales totalled
US$7,000 (not shown) from 11 sales: six cohort
sales (Session 5) and five additional community
sales. Of the 42 (77%) women that sold at least
one piece of jewellery, the median value of jewellery
sold was US$112.80 per woman, ranging from
Sources of income comparisons
Table II displays comparisons of participants’
sources of income at baseline and three months
following the intervention. At both time periods, the
women’s three primary sources of income were
state and federal benefits, parents, friends, and
relatives, and trading sex. Between baseline and
follow-up, significant reductions in illicit sources
of income were reported, including: trading sex
for drugs and/or money (100% vs. 71.0%, respec-
tively, pB /0.0005); and selling drugs (35.0% vs.
10.5%, p? /0.021). Women reported a significantly
higher score on the 4-point job self-efficacy index
(2.6 vs. 2.9, p? /0.004), indicating an increase in
their belief in their ability to function and maintain a
Drug utilization and sexual risk behaviour comparisons
Pre- and three-month post intervention comparisons
of recent drug utilization and sexual risk patterns are
displayed in Table III. There were significant reduc-
tions in a number of reported drug use behaviours
prior to, and three months after the intervention:
daily drug use (76.0% versus 55.0%, respectively,
p? /0.0003); daily crack use (27.3% vs. 13.13%,
respectively, p? /0.0140); injection drug use (55.6%
vs. 35.6%, respectively, p? /0.0027); the median
amount of money spent on drugs on a typical day
($52.57 USD vs. $46.71 USD, respectively, p? /
0.001). JEWEL participants also reported significant
Table II. Recent (past three months) sources of income and job
self efficacy score*, baseline and 3-month follow-up, JEWEL
participants, n? /50.
Traded sex for drugs/money
Part of full time licit job
Mean (SD) job self-efficacy
?Categories not mutually exclusive.
*Last three months.
**P-values obtained using Students t-test for matched observa-
tions for continuous data and McNemar’s Test for categorical
Table III. Recent (past three months) drug utilization and sexual risk patterns, pre and 3-month post intervention (n? /50).
Drug Utilization Patterns
Daily drug use
Daily crack use
] /daily injection
never shared needles
Median US$ amount (SD) spent on drugs on a typical day
Median number (SD) of sexual contacts per month
Steady Male Partner
always used condoms during steady partner vaginal sex
Casual Male Partner
always use condoms during casual partner vaginal sex
Sex Trade Partners
always use condoms during trade partner vaginal sex
Median number (SD) of sex trade partners per month
10.0 (29.9)3.0 (22.4)0.010
*P-values obtained using Students t-test for matched observations for continuous data and Mcnemar’s Test for categorical data.
S. G. Sherman et al.
reductions in sexual behaviours before compared to
three months following the intervention, including:
their median number of sexual contacts per month
(10.0 vs. 3.0, respectively, p? /0.010); condom use
during vaginal sex with sex trade partners (53.0% vs.
75.0%, respectively, p? /0.03); and participants’
median number of monthly sex trade partners (9.0
vs. 3.0 respectively, p? /0.025). From baseline to
follow up, 15% of participants reported a higher
number of sex trade partners, 12.5% reported the
same number of sex trade partners, and 72.5%
reported fewer sex trade partners (data not shown).
Univariate and multivariate linear regression models
Univariate and multivariate linear regression models,
locatedin Table IV, examined
of the change in the number of sex trade partners
from baseline to follow up. In the multivariate
model, controlling for the women’s score on the
job self efficacy index and their expenditure on
drugs at baseline, income from the jewelry sale was
associated with a reduction in the number of
sex trade partners at follow-up. Women who received
$100 from jewel sale had eight less trade sex
partners at follow-up compared to baseline. The
results from the simulation analyses verified our
results from the original dataset, indicating the
validity of our results.
The study describes a novel intervention that aimed
to reduce women’s drug and sex- related HIV risk
behaviours as well as increase their belief in their
ability to earn money through licit means by making,
marketing and selling beaded jewellery. To our
knowledge, this pilot study was the first to engage
an active drug using population in such an economic
enhancement project. The intervention’s feasibility
and acceptability was demonstrated by the amount
of sales, the high attendance in the intervention, and
the extent of the women’s participation in the unpaid
open-jewellery making sessions. Through the crea-
tion of an economically enabling environment in
combination with enhancing women’s HIV risk
reduction skills, participants reduced selected sexual
and drug-related HIV risk behaviours. Specifically,
women reported significantly lower rates of non-
injection (i.e. crack smoking) and injection drug use
as well as a reduction in the number of all sexual
partners and sexual trade partners. Compared to the
baseline survey, women also reported significantly
lower rates of illegal sources of income at follow-up.
Additionally, in the presence of other variables,
money earned through jewellery sales was associated
with a significantly lower number of sex trade
partners at follow-up.
One explanation for these reductions is the
combined effects of women learning HIV risk
reduction skills as well as the positive reinforcement
and enhanced belief in their ability to generate licit
employment. Given the evidence supporting the
link between women’s
entry into trading sex for drugs or money, and
diminished capacity for HIV risk reduction, it is
plausible that women’s risk behaviours would de-
crease with realization of alternative economic op-
portunities. While results from the current study
suggest that women’s exposure to the possibility of
gaining licit employment may be related to risk
behaviour change, behaviour change sustainability
is most likely to succeed if women have access to job
training programmes and licit employment opportu-
JEWEL was not intended to be a ‘job training’
programme, rather it aimed to provide women with
income-generating skills and the belief in their ability
to participate in licit employment. Although JEWEL
did not specifically focus on training in general job
Table IV. Univariate and multivariate linear regression model for the change in the number of monthly sex trade partners from baseline to
follow-up, among JEWEL participants (n? /50).
Bivariate ModelsMultivariate Models
Variable Parameter EstimateStandard Error
P-valueParameter Estimate Standard Error
African American (vs. White)
Older age (] /39 years of age*)
Ever dropping out of school
Income from jewellery sale ($USD)
Baseline job self-efficacy score*
Baseline drug expenditure ($USD)
Daily drug use at baseline
Injection at baseline
*Centred at mean (2.54).
The JEWEL project
skills, women’s participation in JEWEL increased
their belief in their ability to successfully obtain and
function at a job. Women’s enhanced job self-
efficacy was reflected in the significantly higher score
at follow-up compared to baseline as well as the
significant relationship, in the presence of other
variables, between the score with a lower number
of monthly trade partners at follow-up. This result
would be strengthened if future studies examined the
correlation between job self-efficacy scores and
employment acquisition and maintenance. Although
some of the ‘higher earners’ may have been better
able to supplement other income sources with
JEWEL proceeds, it was not our expectation that
women would be able to survive on the income
generated from the project. Rather, women’s sense
of employment and job training opportunities were
enhanced by JEWEL participation. Although a
higher proportion of women at follow-up compared
to baseline reported having a part or full-time job,
there was not sufficient power to detect a significant
difference. Additionally, a longer follow-up time
period would have provided a more meaningful
portrait of the intervention’s effects on patterns of
both licit and illicit sources of income.
There are multiple benefits that could result from
job training programmes that address the specific
needs of drug users. Both street-recruited and
treatment-based samples of drug users are largely
unemployed (Sanders-Phillips, 2002; Silverman et
al., 2002; Zanis et al., 1994). Within these popula-
tions, women have significantly lower average house-
hold incomes (Brady et al., 1993) and rates of
employment and education (French et al., 1992)
compared to men. Barriers to obtaining or main-
taining employment may be the result of a combina-
tion of drug use and its associated lifestyle, a lack of
education or training, and limited employment
opportunities (Romero Daza et al., 2003; Weeks, et
al., 1998). The majority of research regarding the
role of employment and vocational training (Come-
ford, 1999) among drug users is in the context of
drug treatment settings. By design, active drug users
are ineligible for such programmes, resulting in a
lack of understanding of the effects of job training
participation on drug use.
JEWEL demonstrated that risk reduction could be
augmented by enhancing HIV prevention and risk
reduction skills and enhancing women’s job self
efficacy. Upon study completion, all remaining
supplies and tools were donated to the most success-
ful and enthusiastic JEWEL participant who con-
tinues to hold weekly jewellery-making classes.
Anecdotally, four JEWEL participants and four
new women have continued making and selling
jewellery beyond the study period. The four JEWEL
participants both recruited the new women and
trained them. As of 22 months after the last study-
sponsored sale, they have had a number of sales at
local markets and fairs and have begun to sell pieces
on the internet. This demonstrates the feasibility of
the longer-term viability of such an economic
endeavour with the target population as well as
women’s interest and ability to be involved in licit
income generating opportunities.
Results of the current study must be viewed in
light of several limitations pertaining to both the
study design and intervention. First, the small
number of study participants prevented evaluation
of the intervention’s effects on important but rela-
tively rare behaviours such as condom use with
steady and casual partners and inhibited our ability
to utilize multivariate modelling of drug use pat-
terns. Second, the lack of a control group prevented
our teasing a part the effects of the HIV prevention
and jewellery-making components on behaviour
change. But the data were systematically structured,
signifying the intervention altered predicted factors
rather than solely influencing nonspecific ones such
as socially desirability. In the multivariate model, the
amount of money earned from the intervention was
significantly related to a decreased number of sex
trade partners from baseline to follow-up, suggesting
that jewellery-making component contributed to the
Third, the study’s internal validity is threatening by
social desirability due to the sensitive nature of some
of the survey questions regarding sexual and drug-
related risk behaviours. The combined experienced
of the interviewers combined with the repetition of
several sexual and drug-related questions in the
surveys aimed to minimize this bias. Additionally,
studies on self-reported HIV-risk behaviours of
IDUs have found self-reports of both sexual and
drug use variables to be reliable (Darke, 1998;
Goldstein et al., 1995). Additionally, the three-
month follow-up period does not speak to the
sustainability of behaviour change and significant
results could be partially attributed to regression
towards the mean. A larger study with more parti-
cipants, a more rigorous design, and several follow-
up periods over a longer period of time is needed to
expand the current findings.
We did not examine the complex range of factors
that could be related to women’s involvement in
trading sex. Such intrapersonal factors as early
childhood trauma, psychiatric disorders, and experi-
ences of violence could be directly related to
women’s involvement in both drug use and transac-
tional sex. The survey did not measure and the
intervention did not address these issues that might
be powerful forces in keeping them entrenched in
this deleterious lifestyle. The intervention character-
ized by several limitations. First, although the
S. G. Sherman et al.
intervention aimed to examine the viability of an
HIV prevention and interest in an income-generat-
ing endeavour, the women’s long-term economic
needs were not specifically addressed.
Economic enhancement can enhance the lives of
urban, drug using women involved in trading sex in a
myriad of ways. The current study, though modest in
size, demonstrates the potential for enhancing wo-
men’s health behaviours, their belief in their abilities
to generate licit income, and their self efficacy in
obtaining and maintaining employment. JEWEL’s
success, as evidenced both in the participants’
involvement as well as its affects on sexual risk
taking and drug utilization patterns shows the
feasibility of interventions that aim to change the
structure of women’s lives. Through intercepting the
‘nexus of risk’ (O’Leary, 2001) that both directly and
indirectly shape women’s risk behaviour, behaviour
change sustainability could be achieved.
This pilot study is a first step in developing a
comprehensive HIV prevention intervention that
addresses economic disadvantage as an HIV risk
factor. Intervening on this level not only serves the
public health aim of reducing HIV in this popula-
tion, but addresses low income populations’desire to
ameliorate their economic circumstances.
The study was supported by a grant from NIDA
(DA88110). We would like to thank Eddy Poole for
his commitment to the project, as well as the JEWEL
participants who shared their creativity with us.
Arria, A.M. (2003). Drug treatment completion and post-
discharge employment in the TOPPS-II Interstate Cooperative
Study. Journal of Substance Abuse Treatment, 25(1), 9?/18.
Bandura, A. (2001). Social cognitive theory: An agentic perspec-
tive. Annual Review of Psychology, 52, 1?/26.
Bandura, A. (1986). Social Foundations of Thought and Action: A
Social Cognitive Theory. Upper Saddle River, NJ: Prentice Hall.
Bandura, A. (1989). Human agency in social cognitive theory.
American Psychologist, 44(9), 1175?/1184.
Bandura, A. (1977). A Social Learning Theory. Englewood Hill,
Baseman, J., Ross, M., & Williams, M. (1999). Sale of sex for
drugs and drugs for sex: An economic context of sexual risk
behavior for STDs. Sexually Transmitted Diseases, 26(8), 444?/
Beeker, C., Guenther-Grey, C., & Raj, A. (1998). Community
empowerment paradigm drift and the primary prevention of
HIV/AIDS. Social Science and Medicine, 46(7), 831?/842.
Bigelow, G., & Silverman, K. (1999). Theoretical and empirical
foundations of contingency management treatments for drug
abuse. In S. Higgens, & K. Silverman (Eds.), Motivating
behavior change among illicit-drug abusers: Research on contingency
management interventions (pp. 5?/31). Washington, DC: Amer-
ican Psychological Association.
Brady, K., Grice, D., Dustan, L., & Randall, C. (1993). Gender
differences in substance use disorders. American Journal of
Psychiatry, 150(11), 1707?/1711.
Brown, E. (1991). Community action for health promotion: A
strategy to empower individuals and communities. International
Journal of Health Services, 21(3), 441?/456.
Carmine, E.G., & Zeller, R.A. (1979). Reliability and Validity
Assessment. Beverly Hills, CA: Sage Publications.
Centers for Disease Control and Prevention (2003). Sexually
Transmitted Disease Surveillance Report 2002. Atlanta, GA: U.S.
Department of Health and Human Services.
Chiasson, M.A., Stoneburner, R.L., Hildebrandt, D.S., Ewing,
W.E., Telzak, E.E., & Jaffe, H.W. (1991). Heterosexual
transmission of HIV-1 associated with the use of smokable
freebase cocaine (crack). AIDS, 5(9), 1121?/1126.
Comerford, A.W. (1999). Work dysfunction and addiction.
Common roots. Journal of Substance Abuse Treatment, 16(3),
Connors, M. (1996). Sex, drugs, and structural violence: Un-
raveling the epidemic among poor women in the United States.
In P. Farmer, M. Connors, & J. Simmons (Eds.), Women,
Poverty, and AIDS: Sex, Drugs, and Structural Violence (pp. 91?/
124). Monroe, Maine: Common Courage Press.
Dehovitz, J., Kelly, P., Feldman, J., Sierra, M., Clarke, L.,
Bromberg, J., et al. (1994). Sexually transmitted diseases,
sexual behavior, and cocaine use in inner-city women. American
Journal of Epidemiolology, 140(12), 1125?/1134.
Des Jarlais, D., Abdul-Quader, A., & Tross, S. (1991). The next
problem: Maintenance of AIDS risk reduction among intrave-
nous drug users. International Journal of the Addictions, 26(12),
Doherty, M.C., Garfein, R.S., Monterroso, E., Brown, D., &
Vlahov, D. (2000). Correlates of HIV infection among young
adult short-term injection drug users. AIDS, 14(6), 717?/726.
Farmer, P., Connors, M., & Simmons, J. (1996). Sex, drugs and
structural violence. In Women, Poverty, and AIDS: Sex, Drugs,
and Structural Violence. Monroe, ME: Common Courage Press.
Fenton, L. (2004). Preventing HIV/AIDS through poverty
reduction: the only sustainable solution? Lancet, 364, 1186?/
Feucht, T., Stephens, R., & Sullivan, T. (1993). Drug use patterns
among injection drug users and their sex partners. In Handbook
on risk of AIDS: Injection drug users and sexual partners (pp. 91?/
115). Westport, CN: Greenwood Press.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and
behavior: An introduction to theory and research. Reading, MA:
French, M.T., Dennis, M.L., Mcdougal, G.L., Karuntzos, G.T.,
& Hubbard, R.L. (1992). Training and employment programs
in methadone treatment: Client needs and desires. Journal of
Substance Abuse Treatment, 9(4), 293?/303.
Friedman, S.R. (1998). HIV-related politics in long-term per-
spective. AIDS Care, 10(Suppl 2), 93?/103.
Gawin, F.H., & Ellinwood, E.H., Jr. (1988). Cocaine and other
stimulants. Actions, abuse, and treatment. New England Journal
of Medicine, 318(18), 1173?/1182.
Hashemi, S., Schuler, S., & Riley, A. (1996). Rural credit
programs and women’s empowerment in Bangladesh. World
Development, 24(4), 635?/663.
Higgins, S.T., & Silverman, K. (1999). Motivating behaviors change
among illicit-drug abusers: Research on contingency management
interventions. Washington, D.C.: American Psychological Asso-
Kaiser, H.F. (1960). The application of electronic computers to
factor analysis. Educational and Psychological Measurement, 20,
The JEWEL project
Kelly, J., ST. Lawrence, J., Hood, H., & ST. Lawrence, J. (1989).
Behavioral interventions to reduce AIDS risk activities. Journal
of Consulting Clinical Psychology, 57(1), 60?/7.
Klein, C., Easton, D., & Parker, R. (2002). Structural barriers and
facilitators in HIV prevention: A review of international
research. In A. O’Leary (Ed.), Beyond Condoms: Alternate
Approaches to HIV Prevention (pp. 17?/46). New York, NY:
Kluwer Academic Plenum Publishers.
Kohler, U. (2002). Personal Communication.
Kral, A.H., Lorvick, J., & Edlin, B.R. (2000). Sex- and drug-
related risk among populations of younger and older injection
drug users in adjacent neighborhoods in San Francisco. Journal
of Acquired Immune Deficiency Syndrome, 24(2), 162?/167.
Latka, M., Ahern, J., Garfein, R.S., Ouellet, L., Kerndt, P.,
Morse, P., et al. (2001). Prevalence, incidence, and correlates
of chlamydia and gonorrhea among young adult injection drug
users. Journal of Substance Abuse, 13(1?/2), 73?/88.
Latkin, C.A., Sherman, S.G., & Knowlton, A. (2003). HIV
prevention among drug users: Outcome of a network-oriented
peer outreach intervention. Health Psychology, 22, 332?/339.
Latkin, C., & Knowlton, A. (2000). New directions in HIV
prevention among drug users: Settings, norms, and network
approaches to AIDS prevention (SNNAAP): A social influence
approach. Advances in Medical Sociology, 7, 261?/287.
Milby, J.B., Schumacher, J.E., McNamara, C., Wallace, D.,
Usdan, S., McGill, T., et al. (2000). Initiating abstinence in
cocaine abusing dually diagnosed homeless persons. Drug and
Alcohol Dependence, 60(1), 55?/67.
Milby, J.B., Schumacher, J.E., Wallace, D., Frison, S., McNa-
mara, C., Usdan, S., et al. (2003). Day treatment with
contingency management for cocaine abuse in homeless
persons: 12-month follow-up. Journal of Consulting Clinical
Psychology, 71(3), 619?/621.
Miller, M., & Neaigus, A. (2001). Networks, resources and risk
among women who use drugs. Social Science and Medicine,
Montano, D., Kasprzyk, D., & Taplin, S. (1997). The theory of
reasoned action and the theory of planned behavior. In K.
Glanz, F.M. Lewis, & B.K. Rimer (Eds.), Health Behavior and
Health Education. San Francisco, CA: Jossey-Bass Publishers.
Nelson, K.E., Vlahov, D., Solomon, L., Cohn, S., & Munoz, A.
(1995). Temporal trends of incident human immunodeficiency
virus infection in a cohort of injecting drug users in Baltimore,
Md. Archives of Internal Medicine, 155(12), 1305?/1311.
Nimh Multisite HIV Prevention Trial (1997). Quality control and
quality assurance in HIV prevention research: model from a
multisite HIV prevention trial. AIDS, 11, S4953.
Oakley, A., Fullerton, D., & Holland, J. (1995). Behavioral
interventions for HIV/AIDS prevention. AIDS, 9(5), 479?/
O’Leary, A. (2001). Substance use and HIV. Disentangling the
nexus of risk. Journal of Substance Abuse, 1?/2(13), 1?/3.
Platt, J.J. (1995). Vocational rehabilitation of drug abusers.
Psychological Bulletin, 117(3), 416?/433.
Rhodes, T., & Quirk, A. (1998). Drug users’ sexual relationships
and the social organisation of risk: The sexual relationship as a
site of risk management. Social Science and Medicine, 46(2),
Rhodes, T., Stimson, G., & Quirk, A. (1996). Sex, drugs,
intervention and research: from the individual to the social.
Substance Use and Misuse, 31(3), 375?/407.
Romero-Daza, N., Weeks, M., & Singer, M. (2003). ‘Nobody
gives a damn if I live or die’: Violence, drugs, and street-level
prostitution in inner-city Hartford, Connecticut. Medical
Anthropology, 22(3), 233?/259.
Ross, M., Hwang, L., Zack, C., Bull, L., & Williams, M. (2002).
Sexual risk behaviours and STIs in drug abuse treatment
populations whose drug of choice is crack cocaine. International
Journal of STD AIDS, 13(11), 769?/774.
Ross, M.W., Hwang, L.Y., Leonard, L., Teng, M., & Duncan, L.
(1999). Sexual behaviour, STDs and drug use in a crack house
population. International Journal of STD AIDS, 10(4), 224?/
Sanders-Phillips, K. (2002). Factors influencing HIV/AIDS in
women of color. Public Health Reports, 117(Suppl 1), S151?/
Schottenfeld, R.S., Pascale, R., & Solkolowski, S. (1992).
Matching services to needs. Vocational services for substance
abusers. Journal of Substance Abuse Treatment, 9(1), 3?/8.
Schuler, S., Hashemi, S., & Riley, A. (1997). The influence of
women’s changing roles and status in Bangladesh’s fertility
transition: Evidence from a study of credit programs and
contraceptive use. World Development, 25 (April), 563?/575.
Sebstad, J., & Chen, G. (1996). Overview of studies on the impact of
microenterprise credit. Washington DC: AIMS project/ USAID.
Sherman, S.G., Plitt S., Sapun, M., Viscid, I.R., Fuller, C., &
Strathdee, S.A. (2003). Associations between STIs and pre-
valent HIV Infection among injection and non-injection young
drug users. Paper presented. American Public Health Association
Annual Meeting: San Francisco, CA.
Silverman, K., Svikis, D., Wong, C.J., Hampton, J., Stitzer, M.L.,
& Bigelow, G.E. (2002). A reinforcement-based therapeutic
workplace for the treatment of drug abuse: Three-year absti-
nence outcomes. Experiences in Clinical Psychopharmacology,
Somlai, A.M., Kelly, J.A., Wagstaff, D.A., & Whitson, D.P.
(1998). Patterns, predictors, and situational contexts of HIV
risk behaviors among homeless men and women. Social Work ,
43, 1998(1), 7?/20.
Stevens, S., Tortu, S., & Coyle, S. (1998). Women drug users and
HIV prevention: Overview of findings and research needs.
Women Health, 27(1?/2), 19?/23.
Susser, E., Valencia, E., Berkman, A., Sohler, N., Conover, S.,
Torres, J., et al. (1998). Human immunodeficiency virus sexual
risk reduction in homeless men with mental illness. Archives of
General Psychiatry, 55(3), 266?/272.
Sweat, M.D., & Denison, J.A. (1998). Reducing HIV incidence in
developing countries with structural and environmental inter-
ventions. AIDS, 9, S251?/S257.
Ulin, P.R. (1992). African women and AIDS: Negotiating
behavior change. Social Science and Medicine, 34(1), 63?/72.
Van Empelen, P., Kok, G., Van Kesteren, N., Van den Borne, B.,
Bos, A., & Schaalma, H.P. (2003). Effective methods to change
sex-risk among drug users: A review of psychosocial interven-
tions. Social Science & Medicine, 57(9), 1593?/1608.
Wechsberg, W., Dennis, M., & Stevens, S. (1998). Cluster analysis
of HIV intervention outcomes among substance-abusing wo-
men. American Journal of Drug Alcohol Abuse, 24(2), 239?/257.
Weeks, M.R., Grier, M., Romero-Daza, N., Puglisi-Vasquez, M.,
& Singer, M. (1998). Streets, drugs, and the economy of sex in
the age of AIDS. Women Health, 27(1?/2), 205?/229.
Weissman, G., Snowder, B., & Young, P. (1990). The relationship
between crack cocaine use and other risk factors among women in a
national AIDS prevention program. Paper presented. VI Inter-
national AIDS Conference: San Francisco, CA.
Widman, M., Lidz, V., Digregorio, G.J., Platt, A.K., Robison, L.,
& Platt, J.J. (2000). Health status of employed and unemployed
methadone patients. Journal of Substance Abuse Treatment,
Worth, D. (1990). Sexual decision making and AIDS: Why
condom promotion among vulnerable women is likely to fail.
Studies in Family Planning, 20(6), 297?/307.
Wuensch, K.L. (2002). Simulating data for a multiple regression.
accessed on Oct. 6th, 2004.
S. G. Sherman et al.
Zanis, D., Metzger, D., & McLellan, A. (1994). Factors associated
with employment among methadone patients. Journal of Sub-
stance Abuse Treatment, 11(5), 443?/447.
Zimmerman, M. (2000). Empowerment theory: Psychological,
organizational, and community levels of analysis. In J. Rappa-
port, & E. Seidman (Eds.), Handbook of Community Psychology
(pp. 43?/63). New York: Kluwer Academic/ Plenum Publishers.
Zimmerman, M.A. (1995). Psychological empowerment: Issues
and illustrations. American Journal of Community Psychology,
The JEWEL project